End of training
Browse files- README.md +3 -1
- all_results.json +11 -11
- eval_nbest_predictions.json +2 -2
- eval_predictions.json +0 -0
- eval_results.json +6 -6
- runs/Dec14_15-25-09_xgpi8/events.out.tfevents.1734174966.xgpi8.2459273.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +1527 -57
README.md
CHANGED
|
@@ -4,6 +4,8 @@ license: llama3.2
|
|
| 4 |
base_model: meta-llama/Llama-3.2-1B
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
model-index:
|
| 8 |
- name: squad_llama_finetuned
|
| 9 |
results: []
|
|
@@ -14,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# squad_llama_finetuned
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 4 |
base_model: meta-llama/Llama-3.2-1B
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
datasets:
|
| 8 |
+
- squad
|
| 9 |
model-index:
|
| 10 |
- name: squad_llama_finetuned
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# squad_llama_finetuned
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on the squad dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_exact_match": 0.
|
| 4 |
-
"eval_f1": 4.
|
| 5 |
-
"eval_runtime":
|
| 6 |
"eval_samples": 10787,
|
| 7 |
-
"eval_samples_per_second":
|
| 8 |
-
"eval_steps_per_second": 5.
|
| 9 |
-
"total_flos":
|
| 10 |
-
"train_loss": 5.
|
| 11 |
-
"train_runtime":
|
| 12 |
"train_samples": 88559,
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second":
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_exact_match": 0.24597918637653737,
|
| 4 |
+
"eval_f1": 4.74210177589907,
|
| 5 |
+
"eval_runtime": 254.7436,
|
| 6 |
"eval_samples": 10787,
|
| 7 |
+
"eval_samples_per_second": 42.345,
|
| 8 |
+
"eval_steps_per_second": 5.296,
|
| 9 |
+
"total_flos": 9.928075328339328e+17,
|
| 10 |
+
"train_loss": 5.44822753134351,
|
| 11 |
+
"train_runtime": 13196.2606,
|
| 12 |
"train_samples": 88559,
|
| 13 |
+
"train_samples_per_second": 33.555,
|
| 14 |
+
"train_steps_per_second": 8.389
|
| 15 |
}
|
eval_nbest_predictions.json
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 sha256:4db1712cf9ff38bd4d87fe274552ab64b77a6e20514746c0035c0b6502a76b83
|
| 3 |
+
size 51772963
|
eval_predictions.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_exact_match": 0.
|
| 4 |
-
"eval_f1": 4.
|
| 5 |
-
"eval_runtime":
|
| 6 |
"eval_samples": 10787,
|
| 7 |
-
"eval_samples_per_second":
|
| 8 |
-
"eval_steps_per_second": 5.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_exact_match": 0.24597918637653737,
|
| 4 |
+
"eval_f1": 4.74210177589907,
|
| 5 |
+
"eval_runtime": 254.7436,
|
| 6 |
"eval_samples": 10787,
|
| 7 |
+
"eval_samples_per_second": 42.345,
|
| 8 |
+
"eval_steps_per_second": 5.296
|
| 9 |
}
|
runs/Dec14_15-25-09_xgpi8/events.out.tfevents.1734174966.xgpi8.2459273.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ec11ab2b8194c8c06eccf82d9ba8f9556cb94f13e1363a3dc8efcf74da14956
|
| 3 |
+
size 418
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 5.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 88559,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second":
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 9.928075328339328e+17,
|
| 4 |
+
"train_loss": 5.44822753134351,
|
| 5 |
+
"train_runtime": 13196.2606,
|
| 6 |
"train_samples": 88559,
|
| 7 |
+
"train_samples_per_second": 33.555,
|
| 8 |
+
"train_steps_per_second": 8.389
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,104 +1,1574 @@
|
|
| 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,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 6.
|
| 16 |
"step": 500
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 1000
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 5.
|
| 30 |
"step": 1500
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 5.
|
| 37 |
"step": 2000
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 5.
|
| 44 |
"step": 2500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch":
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 5.
|
| 51 |
"step": 3000
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch":
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 5.
|
| 58 |
"step": 3500
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch":
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 5.
|
| 65 |
"step": 4000
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch":
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 5.
|
| 72 |
"step": 4500
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch":
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 5.
|
| 79 |
"step": 5000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 5.
|
| 86 |
"step": 5500
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch":
|
| 90 |
-
"
|
| 91 |
-
"
|
| 92 |
-
"
|
| 93 |
-
"
|
| 94 |
-
|
| 95 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
}
|
| 97 |
],
|
| 98 |
"logging_steps": 500,
|
| 99 |
-
"max_steps":
|
| 100 |
"num_input_tokens_seen": 0,
|
| 101 |
-
"num_train_epochs":
|
| 102 |
"save_steps": 500,
|
| 103 |
"stateful_callbacks": {
|
| 104 |
"TrainerControl": {
|
|
@@ -112,8 +1582,8 @@
|
|
| 112 |
"attributes": {}
|
| 113 |
}
|
| 114 |
},
|
| 115 |
-
"total_flos":
|
| 116 |
-
"train_batch_size":
|
| 117 |
"trial_name": null,
|
| 118 |
"trial_params": null
|
| 119 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 110700,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.022583559168925023,
|
| 13 |
+
"grad_norm": 14.624160766601562,
|
| 14 |
+
"learning_rate": 1.99096657633243e-05,
|
| 15 |
+
"loss": 6.0013,
|
| 16 |
"step": 500
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.045167118337850046,
|
| 20 |
+
"grad_norm": 15.735489845275879,
|
| 21 |
+
"learning_rate": 1.9819331526648603e-05,
|
| 22 |
+
"loss": 5.9387,
|
| 23 |
"step": 1000
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.06775067750677506,
|
| 27 |
+
"grad_norm": 14.514668464660645,
|
| 28 |
+
"learning_rate": 1.9728997289972902e-05,
|
| 29 |
+
"loss": 5.847,
|
| 30 |
"step": 1500
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.09033423667570009,
|
| 34 |
+
"grad_norm": 15.38024616241455,
|
| 35 |
+
"learning_rate": 1.96386630532972e-05,
|
| 36 |
+
"loss": 5.8336,
|
| 37 |
"step": 2000
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.11291779584462511,
|
| 41 |
+
"grad_norm": 14.74740982055664,
|
| 42 |
+
"learning_rate": 1.95483288166215e-05,
|
| 43 |
+
"loss": 5.7863,
|
| 44 |
"step": 2500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.13550135501355012,
|
| 48 |
+
"grad_norm": 14.431930541992188,
|
| 49 |
+
"learning_rate": 1.94579945799458e-05,
|
| 50 |
+
"loss": 5.7847,
|
| 51 |
"step": 3000
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.15808491418247517,
|
| 55 |
+
"grad_norm": 14.330245971679688,
|
| 56 |
+
"learning_rate": 1.9367660343270102e-05,
|
| 57 |
+
"loss": 5.7502,
|
| 58 |
"step": 3500
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.18066847335140018,
|
| 62 |
+
"grad_norm": 14.765193939208984,
|
| 63 |
+
"learning_rate": 1.92773261065944e-05,
|
| 64 |
+
"loss": 5.7527,
|
| 65 |
"step": 4000
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.2032520325203252,
|
| 69 |
+
"grad_norm": 14.284146308898926,
|
| 70 |
+
"learning_rate": 1.91869918699187e-05,
|
| 71 |
+
"loss": 5.733,
|
| 72 |
"step": 4500
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.22583559168925021,
|
| 76 |
+
"grad_norm": 15.028379440307617,
|
| 77 |
+
"learning_rate": 1.9096657633243002e-05,
|
| 78 |
+
"loss": 5.7332,
|
| 79 |
"step": 5000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.24841915085817526,
|
| 83 |
+
"grad_norm": 14.660799026489258,
|
| 84 |
+
"learning_rate": 1.90063233965673e-05,
|
| 85 |
+
"loss": 5.7017,
|
| 86 |
"step": 5500
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.27100271002710025,
|
| 90 |
+
"grad_norm": 14.359591484069824,
|
| 91 |
+
"learning_rate": 1.89159891598916e-05,
|
| 92 |
+
"loss": 5.6775,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2935862691960253,
|
| 97 |
+
"grad_norm": 14.639114379882812,
|
| 98 |
+
"learning_rate": 1.8825654923215903e-05,
|
| 99 |
+
"loss": 5.6341,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.31616982836495033,
|
| 104 |
+
"grad_norm": 15.092992782592773,
|
| 105 |
+
"learning_rate": 1.87353206865402e-05,
|
| 106 |
+
"loss": 5.6577,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.33875338753387535,
|
| 111 |
+
"grad_norm": 15.00180435180664,
|
| 112 |
+
"learning_rate": 1.86449864498645e-05,
|
| 113 |
+
"loss": 5.6297,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.36133694670280037,
|
| 118 |
+
"grad_norm": 14.052933692932129,
|
| 119 |
+
"learning_rate": 1.85546522131888e-05,
|
| 120 |
+
"loss": 5.6289,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3839205058717254,
|
| 125 |
+
"grad_norm": 14.41851806640625,
|
| 126 |
+
"learning_rate": 1.84643179765131e-05,
|
| 127 |
+
"loss": 5.6009,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4065040650406504,
|
| 132 |
+
"grad_norm": 14.261062622070312,
|
| 133 |
+
"learning_rate": 1.83739837398374e-05,
|
| 134 |
+
"loss": 5.6118,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4290876242095754,
|
| 139 |
+
"grad_norm": 14.190622329711914,
|
| 140 |
+
"learning_rate": 1.82836495031617e-05,
|
| 141 |
+
"loss": 5.6026,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.45167118337850043,
|
| 146 |
+
"grad_norm": 13.900622367858887,
|
| 147 |
+
"learning_rate": 1.8193315266486e-05,
|
| 148 |
+
"loss": 5.5832,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4742547425474255,
|
| 153 |
+
"grad_norm": 15.139060020446777,
|
| 154 |
+
"learning_rate": 1.8102981029810302e-05,
|
| 155 |
+
"loss": 5.5913,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4968383017163505,
|
| 160 |
+
"grad_norm": 14.453042984008789,
|
| 161 |
+
"learning_rate": 1.80126467931346e-05,
|
| 162 |
+
"loss": 5.5562,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5194218608852755,
|
| 167 |
+
"grad_norm": 14.795927047729492,
|
| 168 |
+
"learning_rate": 1.7922312556458897e-05,
|
| 169 |
+
"loss": 5.5631,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5420054200542005,
|
| 174 |
+
"grad_norm": 14.965910911560059,
|
| 175 |
+
"learning_rate": 1.78319783197832e-05,
|
| 176 |
+
"loss": 5.5588,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5645889792231256,
|
| 181 |
+
"grad_norm": 14.482827186584473,
|
| 182 |
+
"learning_rate": 1.7741644083107498e-05,
|
| 183 |
+
"loss": 5.5465,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5871725383920506,
|
| 188 |
+
"grad_norm": 14.006292343139648,
|
| 189 |
+
"learning_rate": 1.7651309846431797e-05,
|
| 190 |
+
"loss": 5.5499,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6097560975609756,
|
| 195 |
+
"grad_norm": 14.196497917175293,
|
| 196 |
+
"learning_rate": 1.75609756097561e-05,
|
| 197 |
+
"loss": 5.564,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6323396567299007,
|
| 202 |
+
"grad_norm": 13.68795394897461,
|
| 203 |
+
"learning_rate": 1.74706413730804e-05,
|
| 204 |
+
"loss": 5.5418,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6549232158988256,
|
| 209 |
+
"grad_norm": 14.1574068069458,
|
| 210 |
+
"learning_rate": 1.7380307136404698e-05,
|
| 211 |
+
"loss": 5.5562,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6775067750677507,
|
| 216 |
+
"grad_norm": 14.386873245239258,
|
| 217 |
+
"learning_rate": 1.7289972899729e-05,
|
| 218 |
+
"loss": 5.532,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7000903342366757,
|
| 223 |
+
"grad_norm": 14.89129638671875,
|
| 224 |
+
"learning_rate": 1.71996386630533e-05,
|
| 225 |
+
"loss": 5.5244,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7226738934056007,
|
| 230 |
+
"grad_norm": 14.07315444946289,
|
| 231 |
+
"learning_rate": 1.71093044263776e-05,
|
| 232 |
+
"loss": 5.546,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7452574525745257,
|
| 237 |
+
"grad_norm": 14.609404563903809,
|
| 238 |
+
"learning_rate": 1.7018970189701897e-05,
|
| 239 |
+
"loss": 5.5192,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7678410117434508,
|
| 244 |
+
"grad_norm": 14.25218391418457,
|
| 245 |
+
"learning_rate": 1.6928635953026196e-05,
|
| 246 |
+
"loss": 5.5127,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7904245709123758,
|
| 251 |
+
"grad_norm": 14.558369636535645,
|
| 252 |
+
"learning_rate": 1.68383017163505e-05,
|
| 253 |
+
"loss": 5.4955,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8130081300813008,
|
| 258 |
+
"grad_norm": 14.221301078796387,
|
| 259 |
+
"learning_rate": 1.6747967479674798e-05,
|
| 260 |
+
"loss": 5.5174,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8355916892502259,
|
| 265 |
+
"grad_norm": 14.1200532913208,
|
| 266 |
+
"learning_rate": 1.6657633242999097e-05,
|
| 267 |
+
"loss": 5.5492,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8581752484191508,
|
| 272 |
+
"grad_norm": 14.250799179077148,
|
| 273 |
+
"learning_rate": 1.65672990063234e-05,
|
| 274 |
+
"loss": 5.5262,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.8807588075880759,
|
| 279 |
+
"grad_norm": 14.543590545654297,
|
| 280 |
+
"learning_rate": 1.64769647696477e-05,
|
| 281 |
+
"loss": 5.4892,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9033423667570009,
|
| 286 |
+
"grad_norm": 14.202034950256348,
|
| 287 |
+
"learning_rate": 1.6386630532971998e-05,
|
| 288 |
+
"loss": 5.4926,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9259259259259259,
|
| 293 |
+
"grad_norm": 13.623809814453125,
|
| 294 |
+
"learning_rate": 1.6296296296296297e-05,
|
| 295 |
+
"loss": 5.5,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.948509485094851,
|
| 300 |
+
"grad_norm": 14.561544418334961,
|
| 301 |
+
"learning_rate": 1.6205962059620596e-05,
|
| 302 |
+
"loss": 5.5177,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.971093044263776,
|
| 307 |
+
"grad_norm": 14.286497116088867,
|
| 308 |
+
"learning_rate": 1.6115627822944898e-05,
|
| 309 |
+
"loss": 5.4939,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.993676603432701,
|
| 314 |
+
"grad_norm": 14.363648414611816,
|
| 315 |
+
"learning_rate": 1.6025293586269197e-05,
|
| 316 |
+
"loss": 5.5012,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.016260162601626,
|
| 321 |
+
"grad_norm": 14.406315803527832,
|
| 322 |
+
"learning_rate": 1.5934959349593496e-05,
|
| 323 |
+
"loss": 5.4621,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.038843721770551,
|
| 328 |
+
"grad_norm": 13.818329811096191,
|
| 329 |
+
"learning_rate": 1.58446251129178e-05,
|
| 330 |
+
"loss": 5.4521,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.061427280939476,
|
| 335 |
+
"grad_norm": 13.946843147277832,
|
| 336 |
+
"learning_rate": 1.5754290876242098e-05,
|
| 337 |
+
"loss": 5.4545,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.084010840108401,
|
| 342 |
+
"grad_norm": 13.875289916992188,
|
| 343 |
+
"learning_rate": 1.5663956639566397e-05,
|
| 344 |
+
"loss": 5.4832,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.1065943992773262,
|
| 349 |
+
"grad_norm": 15.377466201782227,
|
| 350 |
+
"learning_rate": 1.55736224028907e-05,
|
| 351 |
+
"loss": 5.4584,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.1291779584462511,
|
| 356 |
+
"grad_norm": 15.088509559631348,
|
| 357 |
+
"learning_rate": 1.5483288166214995e-05,
|
| 358 |
+
"loss": 5.4561,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.151761517615176,
|
| 363 |
+
"grad_norm": 14.345353126525879,
|
| 364 |
+
"learning_rate": 1.5392953929539297e-05,
|
| 365 |
+
"loss": 5.4557,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.174345076784101,
|
| 370 |
+
"grad_norm": 14.212705612182617,
|
| 371 |
+
"learning_rate": 1.5302619692863596e-05,
|
| 372 |
+
"loss": 5.4625,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.1969286359530262,
|
| 377 |
+
"grad_norm": 14.184494972229004,
|
| 378 |
+
"learning_rate": 1.5212285456187897e-05,
|
| 379 |
+
"loss": 5.4829,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.2195121951219512,
|
| 384 |
+
"grad_norm": 14.079368591308594,
|
| 385 |
+
"learning_rate": 1.5121951219512196e-05,
|
| 386 |
+
"loss": 5.4473,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.2420957542908762,
|
| 391 |
+
"grad_norm": 14.130334854125977,
|
| 392 |
+
"learning_rate": 1.5031616982836497e-05,
|
| 393 |
+
"loss": 5.4705,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.2646793134598013,
|
| 398 |
+
"grad_norm": 13.99329948425293,
|
| 399 |
+
"learning_rate": 1.4941282746160796e-05,
|
| 400 |
+
"loss": 5.4579,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.2872628726287263,
|
| 405 |
+
"grad_norm": 14.875615119934082,
|
| 406 |
+
"learning_rate": 1.4850948509485097e-05,
|
| 407 |
+
"loss": 5.4405,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.3098464317976513,
|
| 412 |
+
"grad_norm": 13.665042877197266,
|
| 413 |
+
"learning_rate": 1.4760614272809397e-05,
|
| 414 |
+
"loss": 5.4441,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.3324299909665762,
|
| 419 |
+
"grad_norm": 14.112915992736816,
|
| 420 |
+
"learning_rate": 1.4670280036133695e-05,
|
| 421 |
+
"loss": 5.4477,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.3550135501355014,
|
| 426 |
+
"grad_norm": 14.36024284362793,
|
| 427 |
+
"learning_rate": 1.4579945799457996e-05,
|
| 428 |
+
"loss": 5.4558,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.3775971093044264,
|
| 433 |
+
"grad_norm": 14.128899574279785,
|
| 434 |
+
"learning_rate": 1.4489611562782295e-05,
|
| 435 |
+
"loss": 5.4399,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.4001806684733513,
|
| 440 |
+
"grad_norm": 14.880925178527832,
|
| 441 |
+
"learning_rate": 1.4399277326106595e-05,
|
| 442 |
+
"loss": 5.4671,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.4227642276422765,
|
| 447 |
+
"grad_norm": 14.057498931884766,
|
| 448 |
+
"learning_rate": 1.4308943089430896e-05,
|
| 449 |
+
"loss": 5.4433,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.4453477868112015,
|
| 454 |
+
"grad_norm": 13.734399795532227,
|
| 455 |
+
"learning_rate": 1.4218608852755195e-05,
|
| 456 |
+
"loss": 5.4539,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.4679313459801264,
|
| 461 |
+
"grad_norm": 13.91814136505127,
|
| 462 |
+
"learning_rate": 1.4128274616079496e-05,
|
| 463 |
+
"loss": 5.439,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.4905149051490514,
|
| 468 |
+
"grad_norm": 15.35647964477539,
|
| 469 |
+
"learning_rate": 1.4037940379403797e-05,
|
| 470 |
+
"loss": 5.4181,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.5130984643179766,
|
| 475 |
+
"grad_norm": 15.262383460998535,
|
| 476 |
+
"learning_rate": 1.3947606142728094e-05,
|
| 477 |
+
"loss": 5.4409,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.5356820234869015,
|
| 482 |
+
"grad_norm": 14.099081993103027,
|
| 483 |
+
"learning_rate": 1.3857271906052395e-05,
|
| 484 |
+
"loss": 5.4261,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.5582655826558267,
|
| 489 |
+
"grad_norm": 14.662623405456543,
|
| 490 |
+
"learning_rate": 1.3766937669376694e-05,
|
| 491 |
+
"loss": 5.4612,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.5808491418247517,
|
| 496 |
+
"grad_norm": 14.23056697845459,
|
| 497 |
+
"learning_rate": 1.3676603432700995e-05,
|
| 498 |
+
"loss": 5.4432,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.6034327009936766,
|
| 503 |
+
"grad_norm": 15.749434471130371,
|
| 504 |
+
"learning_rate": 1.3586269196025295e-05,
|
| 505 |
+
"loss": 5.4394,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6260162601626016,
|
| 510 |
+
"grad_norm": 13.69320011138916,
|
| 511 |
+
"learning_rate": 1.3495934959349594e-05,
|
| 512 |
+
"loss": 5.4402,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.6485998193315266,
|
| 517 |
+
"grad_norm": 13.9033842086792,
|
| 518 |
+
"learning_rate": 1.3405600722673895e-05,
|
| 519 |
+
"loss": 5.4422,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.6711833785004515,
|
| 524 |
+
"grad_norm": 14.946853637695312,
|
| 525 |
+
"learning_rate": 1.3315266485998196e-05,
|
| 526 |
+
"loss": 5.4251,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.6937669376693767,
|
| 531 |
+
"grad_norm": 14.465194702148438,
|
| 532 |
+
"learning_rate": 1.3224932249322495e-05,
|
| 533 |
+
"loss": 5.4222,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.7163504968383019,
|
| 538 |
+
"grad_norm": 14.499493598937988,
|
| 539 |
+
"learning_rate": 1.3134598012646794e-05,
|
| 540 |
+
"loss": 5.4116,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.7389340560072268,
|
| 545 |
+
"grad_norm": 14.03498649597168,
|
| 546 |
+
"learning_rate": 1.3044263775971093e-05,
|
| 547 |
+
"loss": 5.4472,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.7615176151761518,
|
| 552 |
+
"grad_norm": 14.21508502960205,
|
| 553 |
+
"learning_rate": 1.2953929539295394e-05,
|
| 554 |
+
"loss": 5.4475,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.7841011743450768,
|
| 559 |
+
"grad_norm": 13.943710327148438,
|
| 560 |
+
"learning_rate": 1.2863595302619693e-05,
|
| 561 |
+
"loss": 5.4027,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.8066847335140017,
|
| 566 |
+
"grad_norm": 14.971059799194336,
|
| 567 |
+
"learning_rate": 1.2773261065943994e-05,
|
| 568 |
+
"loss": 5.4088,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.8292682926829267,
|
| 573 |
+
"grad_norm": 14.763011932373047,
|
| 574 |
+
"learning_rate": 1.2682926829268294e-05,
|
| 575 |
+
"loss": 5.4261,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.8518518518518519,
|
| 580 |
+
"grad_norm": 13.903670310974121,
|
| 581 |
+
"learning_rate": 1.2592592592592593e-05,
|
| 582 |
+
"loss": 5.4293,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.8744354110207768,
|
| 587 |
+
"grad_norm": 14.187458992004395,
|
| 588 |
+
"learning_rate": 1.2502258355916894e-05,
|
| 589 |
+
"loss": 5.4074,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.897018970189702,
|
| 594 |
+
"grad_norm": 14.419677734375,
|
| 595 |
+
"learning_rate": 1.2411924119241195e-05,
|
| 596 |
+
"loss": 5.4159,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.919602529358627,
|
| 601 |
+
"grad_norm": 14.214756965637207,
|
| 602 |
+
"learning_rate": 1.2321589882565492e-05,
|
| 603 |
+
"loss": 5.458,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.942186088527552,
|
| 608 |
+
"grad_norm": 14.710817337036133,
|
| 609 |
+
"learning_rate": 1.2231255645889793e-05,
|
| 610 |
+
"loss": 5.4192,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.9647696476964769,
|
| 615 |
+
"grad_norm": 14.859943389892578,
|
| 616 |
+
"learning_rate": 1.2140921409214092e-05,
|
| 617 |
+
"loss": 5.4501,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.9873532068654018,
|
| 622 |
+
"grad_norm": 14.313403129577637,
|
| 623 |
+
"learning_rate": 1.2050587172538393e-05,
|
| 624 |
+
"loss": 5.4185,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.009936766034327,
|
| 629 |
+
"grad_norm": 14.823176383972168,
|
| 630 |
+
"learning_rate": 1.1960252935862693e-05,
|
| 631 |
+
"loss": 5.4214,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.032520325203252,
|
| 636 |
+
"grad_norm": 15.034602165222168,
|
| 637 |
+
"learning_rate": 1.1869918699186992e-05,
|
| 638 |
+
"loss": 5.4081,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.055103884372177,
|
| 643 |
+
"grad_norm": 14.684073448181152,
|
| 644 |
+
"learning_rate": 1.1779584462511293e-05,
|
| 645 |
+
"loss": 5.4052,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.077687443541102,
|
| 650 |
+
"grad_norm": 13.799320220947266,
|
| 651 |
+
"learning_rate": 1.1689250225835594e-05,
|
| 652 |
+
"loss": 5.3956,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.100271002710027,
|
| 657 |
+
"grad_norm": 13.712013244628906,
|
| 658 |
+
"learning_rate": 1.1598915989159893e-05,
|
| 659 |
+
"loss": 5.4038,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.122854561878952,
|
| 664 |
+
"grad_norm": 14.247920036315918,
|
| 665 |
+
"learning_rate": 1.1508581752484192e-05,
|
| 666 |
+
"loss": 5.398,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.145438121047877,
|
| 671 |
+
"grad_norm": 13.575353622436523,
|
| 672 |
+
"learning_rate": 1.1418247515808491e-05,
|
| 673 |
+
"loss": 5.4226,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.168021680216802,
|
| 678 |
+
"grad_norm": 13.822168350219727,
|
| 679 |
+
"learning_rate": 1.1327913279132792e-05,
|
| 680 |
+
"loss": 5.3973,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.1906052393857274,
|
| 685 |
+
"grad_norm": 13.903643608093262,
|
| 686 |
+
"learning_rate": 1.1237579042457093e-05,
|
| 687 |
+
"loss": 5.4345,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.2131887985546523,
|
| 692 |
+
"grad_norm": 13.772014617919922,
|
| 693 |
+
"learning_rate": 1.1147244805781392e-05,
|
| 694 |
+
"loss": 5.4186,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.2357723577235773,
|
| 699 |
+
"grad_norm": 13.997783660888672,
|
| 700 |
+
"learning_rate": 1.1056910569105692e-05,
|
| 701 |
+
"loss": 5.3882,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.2583559168925023,
|
| 706 |
+
"grad_norm": 13.945893287658691,
|
| 707 |
+
"learning_rate": 1.0966576332429991e-05,
|
| 708 |
+
"loss": 5.4137,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.280939476061427,
|
| 713 |
+
"grad_norm": 14.775089263916016,
|
| 714 |
+
"learning_rate": 1.0876242095754292e-05,
|
| 715 |
+
"loss": 5.3964,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.303523035230352,
|
| 720 |
+
"grad_norm": 13.861775398254395,
|
| 721 |
+
"learning_rate": 1.0785907859078593e-05,
|
| 722 |
+
"loss": 5.3966,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.326106594399277,
|
| 727 |
+
"grad_norm": 14.2053804397583,
|
| 728 |
+
"learning_rate": 1.069557362240289e-05,
|
| 729 |
+
"loss": 5.4203,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.348690153568202,
|
| 734 |
+
"grad_norm": 14.53366756439209,
|
| 735 |
+
"learning_rate": 1.0605239385727191e-05,
|
| 736 |
+
"loss": 5.4021,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.3712737127371275,
|
| 741 |
+
"grad_norm": 14.52110767364502,
|
| 742 |
+
"learning_rate": 1.051490514905149e-05,
|
| 743 |
+
"loss": 5.3939,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.3938572719060525,
|
| 748 |
+
"grad_norm": 14.300971031188965,
|
| 749 |
+
"learning_rate": 1.0424570912375791e-05,
|
| 750 |
+
"loss": 5.402,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.4164408310749774,
|
| 755 |
+
"grad_norm": 14.447431564331055,
|
| 756 |
+
"learning_rate": 1.0334236675700092e-05,
|
| 757 |
+
"loss": 5.4278,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.4390243902439024,
|
| 762 |
+
"grad_norm": 14.584935188293457,
|
| 763 |
+
"learning_rate": 1.024390243902439e-05,
|
| 764 |
+
"loss": 5.4012,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.4616079494128273,
|
| 769 |
+
"grad_norm": 14.45771312713623,
|
| 770 |
+
"learning_rate": 1.0153568202348691e-05,
|
| 771 |
+
"loss": 5.3919,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.4841915085817523,
|
| 776 |
+
"grad_norm": 14.170404434204102,
|
| 777 |
+
"learning_rate": 1.0063233965672992e-05,
|
| 778 |
+
"loss": 5.3963,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.5067750677506773,
|
| 783 |
+
"grad_norm": 13.865776062011719,
|
| 784 |
+
"learning_rate": 9.972899728997291e-06,
|
| 785 |
+
"loss": 5.4135,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.5293586269196027,
|
| 790 |
+
"grad_norm": 14.329898834228516,
|
| 791 |
+
"learning_rate": 9.88256549232159e-06,
|
| 792 |
+
"loss": 5.4031,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.5519421860885276,
|
| 797 |
+
"grad_norm": 13.990058898925781,
|
| 798 |
+
"learning_rate": 9.792231255645891e-06,
|
| 799 |
+
"loss": 5.4214,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.5745257452574526,
|
| 804 |
+
"grad_norm": 13.485710144042969,
|
| 805 |
+
"learning_rate": 9.70189701897019e-06,
|
| 806 |
+
"loss": 5.3874,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.5971093044263776,
|
| 811 |
+
"grad_norm": 14.275350570678711,
|
| 812 |
+
"learning_rate": 9.61156278229449e-06,
|
| 813 |
+
"loss": 5.3972,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.6196928635953025,
|
| 818 |
+
"grad_norm": 14.40979290008545,
|
| 819 |
+
"learning_rate": 9.52122854561879e-06,
|
| 820 |
+
"loss": 5.403,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.642276422764228,
|
| 825 |
+
"grad_norm": 14.07226276397705,
|
| 826 |
+
"learning_rate": 9.43089430894309e-06,
|
| 827 |
+
"loss": 5.3918,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.6648599819331524,
|
| 832 |
+
"grad_norm": 13.841882705688477,
|
| 833 |
+
"learning_rate": 9.340560072267391e-06,
|
| 834 |
+
"loss": 5.4008,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.687443541102078,
|
| 839 |
+
"grad_norm": 13.820773124694824,
|
| 840 |
+
"learning_rate": 9.25022583559169e-06,
|
| 841 |
+
"loss": 5.4077,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.710027100271003,
|
| 846 |
+
"grad_norm": 13.4264554977417,
|
| 847 |
+
"learning_rate": 9.15989159891599e-06,
|
| 848 |
+
"loss": 5.393,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.7326106594399278,
|
| 853 |
+
"grad_norm": 13.905360221862793,
|
| 854 |
+
"learning_rate": 9.06955736224029e-06,
|
| 855 |
+
"loss": 5.3995,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.7551942186088527,
|
| 860 |
+
"grad_norm": 14.013806343078613,
|
| 861 |
+
"learning_rate": 8.97922312556459e-06,
|
| 862 |
+
"loss": 5.3987,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.7777777777777777,
|
| 867 |
+
"grad_norm": 14.268304824829102,
|
| 868 |
+
"learning_rate": 8.888888888888888e-06,
|
| 869 |
+
"loss": 5.4045,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.8003613369467026,
|
| 874 |
+
"grad_norm": 13.507598876953125,
|
| 875 |
+
"learning_rate": 8.798554652213189e-06,
|
| 876 |
+
"loss": 5.406,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.8229448961156276,
|
| 881 |
+
"grad_norm": 13.630600929260254,
|
| 882 |
+
"learning_rate": 8.70822041553749e-06,
|
| 883 |
+
"loss": 5.4065,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.845528455284553,
|
| 888 |
+
"grad_norm": 14.708646774291992,
|
| 889 |
+
"learning_rate": 8.617886178861789e-06,
|
| 890 |
+
"loss": 5.4044,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.868112014453478,
|
| 895 |
+
"grad_norm": 14.174214363098145,
|
| 896 |
+
"learning_rate": 8.52755194218609e-06,
|
| 897 |
+
"loss": 5.382,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.890695573622403,
|
| 902 |
+
"grad_norm": 14.384222030639648,
|
| 903 |
+
"learning_rate": 8.437217705510389e-06,
|
| 904 |
+
"loss": 5.4053,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.913279132791328,
|
| 909 |
+
"grad_norm": 13.988286018371582,
|
| 910 |
+
"learning_rate": 8.34688346883469e-06,
|
| 911 |
+
"loss": 5.4233,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.935862691960253,
|
| 916 |
+
"grad_norm": 15.10951042175293,
|
| 917 |
+
"learning_rate": 8.256549232158988e-06,
|
| 918 |
+
"loss": 5.4259,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.958446251129178,
|
| 923 |
+
"grad_norm": 14.060369491577148,
|
| 924 |
+
"learning_rate": 8.16621499548329e-06,
|
| 925 |
+
"loss": 5.4085,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.9810298102981028,
|
| 930 |
+
"grad_norm": 14.418787002563477,
|
| 931 |
+
"learning_rate": 8.075880758807588e-06,
|
| 932 |
+
"loss": 5.4018,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.003613369467028,
|
| 937 |
+
"grad_norm": 13.697625160217285,
|
| 938 |
+
"learning_rate": 7.985546522131889e-06,
|
| 939 |
+
"loss": 5.4016,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.026196928635953,
|
| 944 |
+
"grad_norm": 13.646669387817383,
|
| 945 |
+
"learning_rate": 7.895212285456188e-06,
|
| 946 |
+
"loss": 5.406,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.048780487804878,
|
| 951 |
+
"grad_norm": 13.908599853515625,
|
| 952 |
+
"learning_rate": 7.804878048780489e-06,
|
| 953 |
+
"loss": 5.4165,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.071364046973803,
|
| 958 |
+
"grad_norm": 14.83110523223877,
|
| 959 |
+
"learning_rate": 7.71454381210479e-06,
|
| 960 |
+
"loss": 5.4057,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.093947606142728,
|
| 965 |
+
"grad_norm": 13.868396759033203,
|
| 966 |
+
"learning_rate": 7.624209575429088e-06,
|
| 967 |
+
"loss": 5.3987,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.116531165311653,
|
| 972 |
+
"grad_norm": 14.185578346252441,
|
| 973 |
+
"learning_rate": 7.5338753387533885e-06,
|
| 974 |
+
"loss": 5.3744,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.139114724480578,
|
| 979 |
+
"grad_norm": 14.473483085632324,
|
| 980 |
+
"learning_rate": 7.443541102077688e-06,
|
| 981 |
+
"loss": 5.3919,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.1616982836495033,
|
| 986 |
+
"grad_norm": 14.068596839904785,
|
| 987 |
+
"learning_rate": 7.353206865401988e-06,
|
| 988 |
+
"loss": 5.3923,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.1842818428184283,
|
| 993 |
+
"grad_norm": 13.663309097290039,
|
| 994 |
+
"learning_rate": 7.262872628726287e-06,
|
| 995 |
+
"loss": 5.3883,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.2068654019873533,
|
| 1000 |
+
"grad_norm": 14.190384864807129,
|
| 1001 |
+
"learning_rate": 7.172538392050587e-06,
|
| 1002 |
+
"loss": 5.3938,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.229448961156278,
|
| 1007 |
+
"grad_norm": 13.847211837768555,
|
| 1008 |
+
"learning_rate": 7.082204155374888e-06,
|
| 1009 |
+
"loss": 5.3757,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.252032520325203,
|
| 1014 |
+
"grad_norm": 14.141094207763672,
|
| 1015 |
+
"learning_rate": 6.991869918699188e-06,
|
| 1016 |
+
"loss": 5.4018,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.274616079494128,
|
| 1021 |
+
"grad_norm": 14.698291778564453,
|
| 1022 |
+
"learning_rate": 6.901535682023488e-06,
|
| 1023 |
+
"loss": 5.3825,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.297199638663053,
|
| 1028 |
+
"grad_norm": 13.939960479736328,
|
| 1029 |
+
"learning_rate": 6.811201445347787e-06,
|
| 1030 |
+
"loss": 5.3876,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.3197831978319785,
|
| 1035 |
+
"grad_norm": 14.23877239227295,
|
| 1036 |
+
"learning_rate": 6.7208672086720876e-06,
|
| 1037 |
+
"loss": 5.4027,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.3423667570009035,
|
| 1042 |
+
"grad_norm": 14.564846992492676,
|
| 1043 |
+
"learning_rate": 6.6305329719963875e-06,
|
| 1044 |
+
"loss": 5.3959,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.3649503161698284,
|
| 1049 |
+
"grad_norm": 13.720281600952148,
|
| 1050 |
+
"learning_rate": 6.540198735320687e-06,
|
| 1051 |
+
"loss": 5.4267,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.3875338753387534,
|
| 1056 |
+
"grad_norm": 15.00568675994873,
|
| 1057 |
+
"learning_rate": 6.449864498644986e-06,
|
| 1058 |
+
"loss": 5.3841,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.4101174345076783,
|
| 1063 |
+
"grad_norm": 14.020421028137207,
|
| 1064 |
+
"learning_rate": 6.359530261969286e-06,
|
| 1065 |
+
"loss": 5.3766,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.4327009936766033,
|
| 1070 |
+
"grad_norm": 13.821393013000488,
|
| 1071 |
+
"learning_rate": 6.269196025293587e-06,
|
| 1072 |
+
"loss": 5.4049,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.4552845528455283,
|
| 1077 |
+
"grad_norm": 13.526660919189453,
|
| 1078 |
+
"learning_rate": 6.178861788617887e-06,
|
| 1079 |
+
"loss": 5.3774,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.4778681120144537,
|
| 1084 |
+
"grad_norm": 15.432093620300293,
|
| 1085 |
+
"learning_rate": 6.088527551942186e-06,
|
| 1086 |
+
"loss": 5.3868,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.5004516711833786,
|
| 1091 |
+
"grad_norm": 13.915629386901855,
|
| 1092 |
+
"learning_rate": 5.998193315266486e-06,
|
| 1093 |
+
"loss": 5.401,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.5230352303523036,
|
| 1098 |
+
"grad_norm": 14.461374282836914,
|
| 1099 |
+
"learning_rate": 5.907859078590787e-06,
|
| 1100 |
+
"loss": 5.3939,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.5456187895212286,
|
| 1105 |
+
"grad_norm": 15.092782974243164,
|
| 1106 |
+
"learning_rate": 5.8175248419150865e-06,
|
| 1107 |
+
"loss": 5.3873,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.5682023486901535,
|
| 1112 |
+
"grad_norm": 14.457600593566895,
|
| 1113 |
+
"learning_rate": 5.7271906052393864e-06,
|
| 1114 |
+
"loss": 5.3699,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.5907859078590785,
|
| 1119 |
+
"grad_norm": 13.92754077911377,
|
| 1120 |
+
"learning_rate": 5.6368563685636855e-06,
|
| 1121 |
+
"loss": 5.4157,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.6133694670280034,
|
| 1126 |
+
"grad_norm": 13.837529182434082,
|
| 1127 |
+
"learning_rate": 5.546522131887986e-06,
|
| 1128 |
+
"loss": 5.3879,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.635953026196929,
|
| 1133 |
+
"grad_norm": 14.759793281555176,
|
| 1134 |
+
"learning_rate": 5.456187895212286e-06,
|
| 1135 |
+
"loss": 5.3965,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.658536585365854,
|
| 1140 |
+
"grad_norm": 14.628218650817871,
|
| 1141 |
+
"learning_rate": 5.365853658536586e-06,
|
| 1142 |
+
"loss": 5.4149,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.6811201445347788,
|
| 1147 |
+
"grad_norm": 14.144693374633789,
|
| 1148 |
+
"learning_rate": 5.275519421860885e-06,
|
| 1149 |
+
"loss": 5.3911,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.7037037037037037,
|
| 1154 |
+
"grad_norm": 14.25294303894043,
|
| 1155 |
+
"learning_rate": 5.185185185185185e-06,
|
| 1156 |
+
"loss": 5.4012,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.7262872628726287,
|
| 1161 |
+
"grad_norm": 13.905450820922852,
|
| 1162 |
+
"learning_rate": 5.094850948509486e-06,
|
| 1163 |
+
"loss": 5.3802,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.7488708220415536,
|
| 1168 |
+
"grad_norm": 15.431106567382812,
|
| 1169 |
+
"learning_rate": 5.004516711833786e-06,
|
| 1170 |
+
"loss": 5.3604,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.7714543812104786,
|
| 1175 |
+
"grad_norm": 15.13500690460205,
|
| 1176 |
+
"learning_rate": 4.9141824751580855e-06,
|
| 1177 |
+
"loss": 5.3888,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.794037940379404,
|
| 1182 |
+
"grad_norm": 13.613816261291504,
|
| 1183 |
+
"learning_rate": 4.823848238482385e-06,
|
| 1184 |
+
"loss": 5.3864,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.816621499548329,
|
| 1189 |
+
"grad_norm": 14.418743133544922,
|
| 1190 |
+
"learning_rate": 4.733514001806685e-06,
|
| 1191 |
+
"loss": 5.3839,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.839205058717254,
|
| 1196 |
+
"grad_norm": 14.12066650390625,
|
| 1197 |
+
"learning_rate": 4.643179765130985e-06,
|
| 1198 |
+
"loss": 5.3619,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.861788617886179,
|
| 1203 |
+
"grad_norm": 14.60769271850586,
|
| 1204 |
+
"learning_rate": 4.552845528455285e-06,
|
| 1205 |
+
"loss": 5.3689,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.884372177055104,
|
| 1210 |
+
"grad_norm": 14.11111068725586,
|
| 1211 |
+
"learning_rate": 4.462511291779585e-06,
|
| 1212 |
+
"loss": 5.3981,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.906955736224029,
|
| 1217 |
+
"grad_norm": 14.375096321105957,
|
| 1218 |
+
"learning_rate": 4.372177055103884e-06,
|
| 1219 |
+
"loss": 5.3723,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.9295392953929538,
|
| 1224 |
+
"grad_norm": 15.005064010620117,
|
| 1225 |
+
"learning_rate": 4.281842818428185e-06,
|
| 1226 |
+
"loss": 5.3904,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.952122854561879,
|
| 1231 |
+
"grad_norm": 14.196097373962402,
|
| 1232 |
+
"learning_rate": 4.191508581752485e-06,
|
| 1233 |
+
"loss": 5.4017,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.974706413730804,
|
| 1238 |
+
"grad_norm": 14.114343643188477,
|
| 1239 |
+
"learning_rate": 4.101174345076785e-06,
|
| 1240 |
+
"loss": 5.391,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.997289972899729,
|
| 1245 |
+
"grad_norm": 13.409917831420898,
|
| 1246 |
+
"learning_rate": 4.0108401084010845e-06,
|
| 1247 |
+
"loss": 5.3917,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.019873532068654,
|
| 1252 |
+
"grad_norm": 14.153375625610352,
|
| 1253 |
+
"learning_rate": 3.920505871725384e-06,
|
| 1254 |
+
"loss": 5.39,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.042457091237579,
|
| 1259 |
+
"grad_norm": 14.038228988647461,
|
| 1260 |
+
"learning_rate": 3.830171635049684e-06,
|
| 1261 |
+
"loss": 5.3887,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.065040650406504,
|
| 1266 |
+
"grad_norm": 14.152128219604492,
|
| 1267 |
+
"learning_rate": 3.7398373983739838e-06,
|
| 1268 |
+
"loss": 5.3694,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.087624209575429,
|
| 1273 |
+
"grad_norm": 14.053643226623535,
|
| 1274 |
+
"learning_rate": 3.649503161698284e-06,
|
| 1275 |
+
"loss": 5.3964,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.110207768744354,
|
| 1280 |
+
"grad_norm": 13.853004455566406,
|
| 1281 |
+
"learning_rate": 3.5591689250225836e-06,
|
| 1282 |
+
"loss": 5.3892,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 4.132791327913279,
|
| 1287 |
+
"grad_norm": 14.095979690551758,
|
| 1288 |
+
"learning_rate": 3.468834688346884e-06,
|
| 1289 |
+
"loss": 5.3768,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.155374887082204,
|
| 1294 |
+
"grad_norm": 13.6284761428833,
|
| 1295 |
+
"learning_rate": 3.378500451671184e-06,
|
| 1296 |
+
"loss": 5.386,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.177958446251129,
|
| 1301 |
+
"grad_norm": 14.282611846923828,
|
| 1302 |
+
"learning_rate": 3.2881662149954837e-06,
|
| 1303 |
+
"loss": 5.3804,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.200542005420054,
|
| 1308 |
+
"grad_norm": 14.314144134521484,
|
| 1309 |
+
"learning_rate": 3.1978319783197836e-06,
|
| 1310 |
+
"loss": 5.3525,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.22312556458898,
|
| 1315 |
+
"grad_norm": 14.040443420410156,
|
| 1316 |
+
"learning_rate": 3.107497741644083e-06,
|
| 1317 |
+
"loss": 5.3922,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 4.245709123757904,
|
| 1322 |
+
"grad_norm": 14.230873107910156,
|
| 1323 |
+
"learning_rate": 3.0171635049683834e-06,
|
| 1324 |
+
"loss": 5.3577,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 4.2682926829268295,
|
| 1329 |
+
"grad_norm": 14.582721710205078,
|
| 1330 |
+
"learning_rate": 2.926829268292683e-06,
|
| 1331 |
+
"loss": 5.3836,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 4.290876242095754,
|
| 1336 |
+
"grad_norm": 14.114388465881348,
|
| 1337 |
+
"learning_rate": 2.836495031616983e-06,
|
| 1338 |
+
"loss": 5.3775,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 4.313459801264679,
|
| 1343 |
+
"grad_norm": 13.766236305236816,
|
| 1344 |
+
"learning_rate": 2.7461607949412827e-06,
|
| 1345 |
+
"loss": 5.3896,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.336043360433604,
|
| 1350 |
+
"grad_norm": 14.358543395996094,
|
| 1351 |
+
"learning_rate": 2.655826558265583e-06,
|
| 1352 |
+
"loss": 5.3949,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 4.358626919602529,
|
| 1357 |
+
"grad_norm": 14.216902732849121,
|
| 1358 |
+
"learning_rate": 2.565492321589883e-06,
|
| 1359 |
+
"loss": 5.3718,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.381210478771455,
|
| 1364 |
+
"grad_norm": 13.605734825134277,
|
| 1365 |
+
"learning_rate": 2.4751580849141828e-06,
|
| 1366 |
+
"loss": 5.3773,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 4.403794037940379,
|
| 1371 |
+
"grad_norm": 14.073831558227539,
|
| 1372 |
+
"learning_rate": 2.3848238482384827e-06,
|
| 1373 |
+
"loss": 5.3681,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.426377597109305,
|
| 1378 |
+
"grad_norm": 15.32926082611084,
|
| 1379 |
+
"learning_rate": 2.2944896115627826e-06,
|
| 1380 |
+
"loss": 5.3678,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 4.448961156278229,
|
| 1385 |
+
"grad_norm": 13.627253532409668,
|
| 1386 |
+
"learning_rate": 2.204155374887082e-06,
|
| 1387 |
+
"loss": 5.3944,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.471544715447155,
|
| 1392 |
+
"grad_norm": 14.351944923400879,
|
| 1393 |
+
"learning_rate": 2.1138211382113824e-06,
|
| 1394 |
+
"loss": 5.409,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 4.494128274616079,
|
| 1399 |
+
"grad_norm": 14.671781539916992,
|
| 1400 |
+
"learning_rate": 2.0234869015356823e-06,
|
| 1401 |
+
"loss": 5.3932,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 4.5167118337850045,
|
| 1406 |
+
"grad_norm": 14.713647842407227,
|
| 1407 |
+
"learning_rate": 1.933152664859982e-06,
|
| 1408 |
+
"loss": 5.3675,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 4.53929539295393,
|
| 1413 |
+
"grad_norm": 13.966901779174805,
|
| 1414 |
+
"learning_rate": 1.842818428184282e-06,
|
| 1415 |
+
"loss": 5.4028,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 4.561878952122854,
|
| 1420 |
+
"grad_norm": 13.730040550231934,
|
| 1421 |
+
"learning_rate": 1.752484191508582e-06,
|
| 1422 |
+
"loss": 5.3981,
|
| 1423 |
+
"step": 101000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 4.58446251129178,
|
| 1427 |
+
"grad_norm": 13.81587028503418,
|
| 1428 |
+
"learning_rate": 1.6621499548328819e-06,
|
| 1429 |
+
"loss": 5.3775,
|
| 1430 |
+
"step": 101500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.607046070460704,
|
| 1434 |
+
"grad_norm": 13.911078453063965,
|
| 1435 |
+
"learning_rate": 1.5718157181571816e-06,
|
| 1436 |
+
"loss": 5.3878,
|
| 1437 |
+
"step": 102000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 4.62962962962963,
|
| 1441 |
+
"grad_norm": 14.198365211486816,
|
| 1442 |
+
"learning_rate": 1.4814814814814815e-06,
|
| 1443 |
+
"loss": 5.4365,
|
| 1444 |
+
"step": 102500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 4.652213188798554,
|
| 1448 |
+
"grad_norm": 13.877766609191895,
|
| 1449 |
+
"learning_rate": 1.3911472448057814e-06,
|
| 1450 |
+
"loss": 5.3769,
|
| 1451 |
+
"step": 103000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.67479674796748,
|
| 1455 |
+
"grad_norm": 13.929753303527832,
|
| 1456 |
+
"learning_rate": 1.3008130081300815e-06,
|
| 1457 |
+
"loss": 5.3867,
|
| 1458 |
+
"step": 103500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 4.697380307136404,
|
| 1462 |
+
"grad_norm": 14.074589729309082,
|
| 1463 |
+
"learning_rate": 1.2104787714543814e-06,
|
| 1464 |
+
"loss": 5.3692,
|
| 1465 |
+
"step": 104000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 4.71996386630533,
|
| 1469 |
+
"grad_norm": 14.109688758850098,
|
| 1470 |
+
"learning_rate": 1.120144534778681e-06,
|
| 1471 |
+
"loss": 5.4105,
|
| 1472 |
+
"step": 104500
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 4.742547425474255,
|
| 1476 |
+
"grad_norm": 14.187965393066406,
|
| 1477 |
+
"learning_rate": 1.0298102981029812e-06,
|
| 1478 |
+
"loss": 5.4004,
|
| 1479 |
+
"step": 105000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 4.7651309846431795,
|
| 1483 |
+
"grad_norm": 14.166157722473145,
|
| 1484 |
+
"learning_rate": 9.394760614272811e-07,
|
| 1485 |
+
"loss": 5.3468,
|
| 1486 |
+
"step": 105500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 4.787714543812105,
|
| 1490 |
+
"grad_norm": 13.908658027648926,
|
| 1491 |
+
"learning_rate": 8.49141824751581e-07,
|
| 1492 |
+
"loss": 5.3687,
|
| 1493 |
+
"step": 106000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 4.8102981029810294,
|
| 1497 |
+
"grad_norm": 13.582141876220703,
|
| 1498 |
+
"learning_rate": 7.588075880758807e-07,
|
| 1499 |
+
"loss": 5.3807,
|
| 1500 |
+
"step": 106500
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 4.832881662149955,
|
| 1504 |
+
"grad_norm": 13.857187271118164,
|
| 1505 |
+
"learning_rate": 6.684733514001806e-07,
|
| 1506 |
+
"loss": 5.4082,
|
| 1507 |
+
"step": 107000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 4.85546522131888,
|
| 1511 |
+
"grad_norm": 14.338404655456543,
|
| 1512 |
+
"learning_rate": 5.781391147244806e-07,
|
| 1513 |
+
"loss": 5.403,
|
| 1514 |
+
"step": 107500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 4.878048780487805,
|
| 1518 |
+
"grad_norm": 13.691380500793457,
|
| 1519 |
+
"learning_rate": 4.878048780487805e-07,
|
| 1520 |
+
"loss": 5.3817,
|
| 1521 |
+
"step": 108000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 4.90063233965673,
|
| 1525 |
+
"grad_norm": 13.800328254699707,
|
| 1526 |
+
"learning_rate": 3.974706413730804e-07,
|
| 1527 |
+
"loss": 5.3676,
|
| 1528 |
+
"step": 108500
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.923215898825655,
|
| 1532 |
+
"grad_norm": 14.050251007080078,
|
| 1533 |
+
"learning_rate": 3.0713640469738035e-07,
|
| 1534 |
+
"loss": 5.3956,
|
| 1535 |
+
"step": 109000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 4.94579945799458,
|
| 1539 |
+
"grad_norm": 13.885753631591797,
|
| 1540 |
+
"learning_rate": 2.1680216802168024e-07,
|
| 1541 |
+
"loss": 5.387,
|
| 1542 |
+
"step": 109500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 4.968383017163505,
|
| 1546 |
+
"grad_norm": 14.097291946411133,
|
| 1547 |
+
"learning_rate": 1.2646793134598014e-07,
|
| 1548 |
+
"loss": 5.3785,
|
| 1549 |
+
"step": 110000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 4.99096657633243,
|
| 1553 |
+
"grad_norm": 15.280513763427734,
|
| 1554 |
+
"learning_rate": 3.613369467028004e-08,
|
| 1555 |
+
"loss": 5.4046,
|
| 1556 |
+
"step": 110500
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 5.0,
|
| 1560 |
+
"step": 110700,
|
| 1561 |
+
"total_flos": 9.928075328339328e+17,
|
| 1562 |
+
"train_loss": 5.44822753134351,
|
| 1563 |
+
"train_runtime": 13196.2606,
|
| 1564 |
+
"train_samples_per_second": 33.555,
|
| 1565 |
+
"train_steps_per_second": 8.389
|
| 1566 |
}
|
| 1567 |
],
|
| 1568 |
"logging_steps": 500,
|
| 1569 |
+
"max_steps": 110700,
|
| 1570 |
"num_input_tokens_seen": 0,
|
| 1571 |
+
"num_train_epochs": 5,
|
| 1572 |
"save_steps": 500,
|
| 1573 |
"stateful_callbacks": {
|
| 1574 |
"TrainerControl": {
|
|
|
|
| 1582 |
"attributes": {}
|
| 1583 |
}
|
| 1584 |
},
|
| 1585 |
+
"total_flos": 9.928075328339328e+17,
|
| 1586 |
+
"train_batch_size": 4,
|
| 1587 |
"trial_name": null,
|
| 1588 |
"trial_params": null
|
| 1589 |
}
|