End of training
Browse files- README.md +2 -1
- all_results.json +16 -0
- train_results.json +16 -0
- trainer_state.json +1774 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen3-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: taskmaster2-1ep
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# taskmaster2-1ep
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen3-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: taskmaster2-1ep
|
|
|
|
| 16 |
|
| 17 |
# taskmaster2-1ep
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the DCAgent/taskmaster2-1ep dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 2.2154428118318754,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 396.3848338503107,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.4620312452316284,
|
| 7 |
+
"mfu_percent": 0.15656839659589225,
|
| 8 |
+
"mfu_percent_theoretical": 28.013062462919482,
|
| 9 |
+
"total_flos": 2.3852276139845222e+17,
|
| 10 |
+
"train_loss": 0.5620896111628052,
|
| 11 |
+
"train_runtime": 6728.981,
|
| 12 |
+
"train_samples_per_second": 7.428,
|
| 13 |
+
"train_steps_per_second": 0.117,
|
| 14 |
+
"valid_targets_mean": 702.6,
|
| 15 |
+
"valid_targets_min": 459
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 2.2154428118318754,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 396.3848338503107,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.4620312452316284,
|
| 7 |
+
"mfu_percent": 0.15656839659589225,
|
| 8 |
+
"mfu_percent_theoretical": 28.013062462919482,
|
| 9 |
+
"total_flos": 2.3852276139845222e+17,
|
| 10 |
+
"train_loss": 0.5620896111628052,
|
| 11 |
+
"train_runtime": 6728.981,
|
| 12 |
+
"train_samples_per_second": 7.428,
|
| 13 |
+
"train_steps_per_second": 0.117,
|
| 14 |
+
"valid_targets_mean": 702.6,
|
| 15 |
+
"valid_targets_min": 459
|
| 16 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1774 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 785,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.032,
|
| 14 |
+
"grad_norm": 13.448105258759517,
|
| 15 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 16 |
+
"loss": 1.2075,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.31800004839897156,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 748.6,
|
| 21 |
+
"valid_targets_min": 403
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.064,
|
| 25 |
+
"grad_norm": 9.982424737572133,
|
| 26 |
+
"learning_rate": 4.556962025316456e-06,
|
| 27 |
+
"loss": 1.1769,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.3006700873374939,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 714.3,
|
| 32 |
+
"valid_targets_min": 468
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.096,
|
| 36 |
+
"grad_norm": 3.8464182382921748,
|
| 37 |
+
"learning_rate": 7.08860759493671e-06,
|
| 38 |
+
"loss": 1.0706,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.2761875092983246,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 833.6,
|
| 43 |
+
"valid_targets_min": 428
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.128,
|
| 47 |
+
"grad_norm": 2.1620562822697003,
|
| 48 |
+
"learning_rate": 9.620253164556963e-06,
|
| 49 |
+
"loss": 1.0013,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.25137242674827576,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 814.9,
|
| 54 |
+
"valid_targets_min": 392
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.16,
|
| 58 |
+
"grad_norm": 1.6337626383709007,
|
| 59 |
+
"learning_rate": 1.2151898734177216e-05,
|
| 60 |
+
"loss": 0.9265,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.2275439351797104,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 728.6,
|
| 65 |
+
"valid_targets_min": 416
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.192,
|
| 69 |
+
"grad_norm": 1.9026845295535284,
|
| 70 |
+
"learning_rate": 1.468354430379747e-05,
|
| 71 |
+
"loss": 0.8832,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.20272783935070038,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 751.1,
|
| 76 |
+
"valid_targets_min": 439
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.224,
|
| 80 |
+
"grad_norm": 1.073091888557403,
|
| 81 |
+
"learning_rate": 1.7215189873417723e-05,
|
| 82 |
+
"loss": 0.8224,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.21370205283164978,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 786.2,
|
| 87 |
+
"valid_targets_min": 463
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.256,
|
| 91 |
+
"grad_norm": 0.8617650331409201,
|
| 92 |
+
"learning_rate": 1.974683544303798e-05,
|
| 93 |
+
"loss": 0.8024,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.1904214322566986,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 843.4,
|
| 98 |
+
"valid_targets_min": 359
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.288,
|
| 102 |
+
"grad_norm": 0.759980239581234,
|
| 103 |
+
"learning_rate": 2.2278481012658228e-05,
|
| 104 |
+
"loss": 0.7507,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.15183305740356445,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 724.6,
|
| 109 |
+
"valid_targets_min": 343
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32,
|
| 113 |
+
"grad_norm": 0.7463412043901649,
|
| 114 |
+
"learning_rate": 2.481012658227848e-05,
|
| 115 |
+
"loss": 0.736,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.1823093295097351,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 826.6,
|
| 120 |
+
"valid_targets_min": 562
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.352,
|
| 124 |
+
"grad_norm": 0.6886141045868305,
|
| 125 |
+
"learning_rate": 2.7341772151898737e-05,
|
| 126 |
+
"loss": 0.7099,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.17409449815750122,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 738.2,
|
| 131 |
+
"valid_targets_min": 387
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.384,
|
| 135 |
+
"grad_norm": 0.6723991947723206,
|
| 136 |
+
"learning_rate": 2.987341772151899e-05,
|
| 137 |
+
"loss": 0.6817,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.1746065616607666,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 817.4,
|
| 142 |
+
"valid_targets_min": 503
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.416,
|
| 146 |
+
"grad_norm": 0.7219762987855783,
|
| 147 |
+
"learning_rate": 3.240506329113924e-05,
|
| 148 |
+
"loss": 0.7042,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.15542849898338318,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 631.1,
|
| 153 |
+
"valid_targets_min": 418
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.448,
|
| 157 |
+
"grad_norm": 0.711945449946034,
|
| 158 |
+
"learning_rate": 3.49367088607595e-05,
|
| 159 |
+
"loss": 0.6683,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.1588052213191986,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 717.1,
|
| 164 |
+
"valid_targets_min": 483
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48,
|
| 168 |
+
"grad_norm": 0.7723131517405141,
|
| 169 |
+
"learning_rate": 3.746835443037975e-05,
|
| 170 |
+
"loss": 0.6602,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.16386288404464722,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 850.6,
|
| 175 |
+
"valid_targets_min": 252
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.512,
|
| 179 |
+
"grad_norm": 0.7455251121043946,
|
| 180 |
+
"learning_rate": 4e-05,
|
| 181 |
+
"loss": 0.6492,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.16289247572422028,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 734.2,
|
| 186 |
+
"valid_targets_min": 385
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.544,
|
| 190 |
+
"grad_norm": 0.6806696789899241,
|
| 191 |
+
"learning_rate": 3.999504991751045e-05,
|
| 192 |
+
"loss": 0.6491,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.14095750451087952,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 764.7,
|
| 197 |
+
"valid_targets_min": 487
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.576,
|
| 201 |
+
"grad_norm": 0.6997408572409954,
|
| 202 |
+
"learning_rate": 3.9980202120373464e-05,
|
| 203 |
+
"loss": 0.6471,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.15331107378005981,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 785.2,
|
| 208 |
+
"valid_targets_min": 492
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.608,
|
| 212 |
+
"grad_norm": 0.7180343376382498,
|
| 213 |
+
"learning_rate": 3.995546395837111e-05,
|
| 214 |
+
"loss": 0.6481,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.17766906321048737,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 781.8,
|
| 219 |
+
"valid_targets_min": 397
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.64,
|
| 223 |
+
"grad_norm": 0.7022835185401193,
|
| 224 |
+
"learning_rate": 3.992084767709763e-05,
|
| 225 |
+
"loss": 0.6207,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.17354968190193176,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 836.8,
|
| 230 |
+
"valid_targets_min": 418
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.672,
|
| 234 |
+
"grad_norm": 0.7367281503709805,
|
| 235 |
+
"learning_rate": 3.987637041189781e-05,
|
| 236 |
+
"loss": 0.6272,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.162840336561203,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 837.1,
|
| 241 |
+
"valid_targets_min": 434
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.704,
|
| 245 |
+
"grad_norm": 0.6596850958225633,
|
| 246 |
+
"learning_rate": 3.982205417938482e-05,
|
| 247 |
+
"loss": 0.6176,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.14747610688209534,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 715.6,
|
| 252 |
+
"valid_targets_min": 485
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.736,
|
| 256 |
+
"grad_norm": 0.6740339776230146,
|
| 257 |
+
"learning_rate": 3.975792586654179e-05,
|
| 258 |
+
"loss": 0.6183,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.15198436379432678,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 681.2,
|
| 263 |
+
"valid_targets_min": 400
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.768,
|
| 267 |
+
"grad_norm": 0.6279891719181242,
|
| 268 |
+
"learning_rate": 3.968401721741259e-05,
|
| 269 |
+
"loss": 0.6138,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.1494234800338745,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 762.9,
|
| 274 |
+
"valid_targets_min": 313
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8,
|
| 278 |
+
"grad_norm": 0.6762014593224054,
|
| 279 |
+
"learning_rate": 3.960036481738819e-05,
|
| 280 |
+
"loss": 0.6152,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.14409850537776947,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 826.2,
|
| 285 |
+
"valid_targets_min": 432
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.832,
|
| 289 |
+
"grad_norm": 0.6951704276499515,
|
| 290 |
+
"learning_rate": 3.950701007509667e-05,
|
| 291 |
+
"loss": 0.6211,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.1555303931236267,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 755.7,
|
| 296 |
+
"valid_targets_min": 496
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.864,
|
| 300 |
+
"grad_norm": 0.6296778796453734,
|
| 301 |
+
"learning_rate": 3.940399920190552e-05,
|
| 302 |
+
"loss": 0.6129,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.1367054283618927,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 662.6,
|
| 307 |
+
"valid_targets_min": 328
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.896,
|
| 311 |
+
"grad_norm": 0.6783518650883749,
|
| 312 |
+
"learning_rate": 3.92913831890467e-05,
|
| 313 |
+
"loss": 0.5886,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.15633609890937805,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 727.8,
|
| 318 |
+
"valid_targets_min": 507
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.928,
|
| 322 |
+
"grad_norm": 0.708146207975277,
|
| 323 |
+
"learning_rate": 3.916921778237556e-05,
|
| 324 |
+
"loss": 0.5948,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.16752688586711884,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 762.7,
|
| 329 |
+
"valid_targets_min": 541
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.96,
|
| 333 |
+
"grad_norm": 0.6593931091646447,
|
| 334 |
+
"learning_rate": 3.903756345477612e-05,
|
| 335 |
+
"loss": 0.5933,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.164844810962677,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 851.9,
|
| 340 |
+
"valid_targets_min": 389
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.992,
|
| 344 |
+
"grad_norm": 0.6510250678307847,
|
| 345 |
+
"learning_rate": 3.889648537622657e-05,
|
| 346 |
+
"loss": 0.5897,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.1499878615140915,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 824.8,
|
| 351 |
+
"valid_targets_min": 424
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0192,
|
| 355 |
+
"grad_norm": 0.6781043016492474,
|
| 356 |
+
"learning_rate": 3.874605338153952e-05,
|
| 357 |
+
"loss": 0.5918,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.1404564082622528,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 745.8,
|
| 362 |
+
"valid_targets_min": 381
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0512,
|
| 366 |
+
"grad_norm": 0.6508626538105656,
|
| 367 |
+
"learning_rate": 3.8586341935793265e-05,
|
| 368 |
+
"loss": 0.5805,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.14573101699352264,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 690.4,
|
| 373 |
+
"valid_targets_min": 426
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0832,
|
| 377 |
+
"grad_norm": 0.6313686061473457,
|
| 378 |
+
"learning_rate": 3.841743009747089e-05,
|
| 379 |
+
"loss": 0.5801,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.13861840963363647,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 763.2,
|
| 384 |
+
"valid_targets_min": 457
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1152,
|
| 388 |
+
"grad_norm": 0.6718637932023319,
|
| 389 |
+
"learning_rate": 3.8239401479325714e-05,
|
| 390 |
+
"loss": 0.5727,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.13479197025299072,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 740.8,
|
| 395 |
+
"valid_targets_min": 325
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1472,
|
| 399 |
+
"grad_norm": 0.6534534284351228,
|
| 400 |
+
"learning_rate": 3.8052344206992276e-05,
|
| 401 |
+
"loss": 0.5688,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.11501454561948776,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 669.1,
|
| 406 |
+
"valid_targets_min": 361
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1792,
|
| 410 |
+
"grad_norm": 0.6298087553175363,
|
| 411 |
+
"learning_rate": 3.7856350875363396e-05,
|
| 412 |
+
"loss": 0.5533,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.12853585183620453,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 759.6,
|
| 417 |
+
"valid_targets_min": 433
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2112,
|
| 421 |
+
"grad_norm": 0.6629648321914637,
|
| 422 |
+
"learning_rate": 3.765151850275497e-05,
|
| 423 |
+
"loss": 0.5718,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.13259121775627136,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 673.8,
|
| 428 |
+
"valid_targets_min": 428
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.2432,
|
| 432 |
+
"grad_norm": 0.6446910197966146,
|
| 433 |
+
"learning_rate": 3.7437948482881104e-05,
|
| 434 |
+
"loss": 0.5721,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.1383952796459198,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 777.4,
|
| 439 |
+
"valid_targets_min": 559
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2752,
|
| 443 |
+
"grad_norm": 0.6715593999996728,
|
| 444 |
+
"learning_rate": 3.721574653466336e-05,
|
| 445 |
+
"loss": 0.5767,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.13751399517059326,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 690.3,
|
| 450 |
+
"valid_targets_min": 362
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3072,
|
| 454 |
+
"grad_norm": 0.6634970200747077,
|
| 455 |
+
"learning_rate": 3.698502264989903e-05,
|
| 456 |
+
"loss": 0.5749,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.14345307648181915,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 795.1,
|
| 461 |
+
"valid_targets_min": 436
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.3392,
|
| 465 |
+
"grad_norm": 0.6852785997622871,
|
| 466 |
+
"learning_rate": 3.674589103881432e-05,
|
| 467 |
+
"loss": 0.5757,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.16555346548557281,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 953.1,
|
| 472 |
+
"valid_targets_min": 495
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3712,
|
| 476 |
+
"grad_norm": 0.7182921119480419,
|
| 477 |
+
"learning_rate": 3.64984700735293e-05,
|
| 478 |
+
"loss": 0.5662,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.13560381531715393,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 657.9,
|
| 483 |
+
"valid_targets_min": 380
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4032,
|
| 487 |
+
"grad_norm": 0.6383915774320499,
|
| 488 |
+
"learning_rate": 3.624288222946273e-05,
|
| 489 |
+
"loss": 0.5598,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.1385164111852646,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 883.5,
|
| 494 |
+
"valid_targets_min": 351
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4352,
|
| 498 |
+
"grad_norm": 0.6197700397838696,
|
| 499 |
+
"learning_rate": 3.597925402470578e-05,
|
| 500 |
+
"loss": 0.5645,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.12843455374240875,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 750.9,
|
| 505 |
+
"valid_targets_min": 434
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4672,
|
| 509 |
+
"grad_norm": 0.5711420210151498,
|
| 510 |
+
"learning_rate": 3.570771595739445e-05,
|
| 511 |
+
"loss": 0.5598,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.14530006051063538,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 942.8,
|
| 516 |
+
"valid_targets_min": 528
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.4992,
|
| 520 |
+
"grad_norm": 0.6210204975257504,
|
| 521 |
+
"learning_rate": 3.5428402441111964e-05,
|
| 522 |
+
"loss": 0.5593,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.1555243879556656,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 887.2,
|
| 527 |
+
"valid_targets_min": 537
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5312000000000001,
|
| 531 |
+
"grad_norm": 0.6647791326711432,
|
| 532 |
+
"learning_rate": 3.5141451738352936e-05,
|
| 533 |
+
"loss": 0.5609,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.13146094977855682,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 744.5,
|
| 538 |
+
"valid_targets_min": 461
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.5632000000000001,
|
| 542 |
+
"grad_norm": 0.6693133474516781,
|
| 543 |
+
"learning_rate": 3.4847005892082266e-05,
|
| 544 |
+
"loss": 0.577,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.1364261358976364,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 682.8,
|
| 549 |
+
"valid_targets_min": 444
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.5952,
|
| 553 |
+
"grad_norm": 0.7006371063676372,
|
| 554 |
+
"learning_rate": 3.454521065542273e-05,
|
| 555 |
+
"loss": 0.5673,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.13232733309268951,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 724.2,
|
| 560 |
+
"valid_targets_min": 468
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.6272,
|
| 564 |
+
"grad_norm": 0.655773431685838,
|
| 565 |
+
"learning_rate": 3.423621541950597e-05,
|
| 566 |
+
"loss": 0.561,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.13282510638237,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 766.9,
|
| 571 |
+
"valid_targets_min": 402
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.6592,
|
| 575 |
+
"grad_norm": 0.6172192241649032,
|
| 576 |
+
"learning_rate": 3.3920173139522664e-05,
|
| 577 |
+
"loss": 0.5643,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.13580721616744995,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 764.2,
|
| 582 |
+
"valid_targets_min": 400
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.6912,
|
| 586 |
+
"grad_norm": 0.6628955705415062,
|
| 587 |
+
"learning_rate": 3.35972402590084e-05,
|
| 588 |
+
"loss": 0.5581,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.1439439356327057,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 715.3,
|
| 593 |
+
"valid_targets_min": 309
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7231999999999998,
|
| 597 |
+
"grad_norm": 0.641312579448422,
|
| 598 |
+
"learning_rate": 3.326757663240291e-05,
|
| 599 |
+
"loss": 0.5624,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.1477782130241394,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 739.2,
|
| 604 |
+
"valid_targets_min": 472
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7551999999999999,
|
| 608 |
+
"grad_norm": 0.6833918688567646,
|
| 609 |
+
"learning_rate": 3.293134544592073e-05,
|
| 610 |
+
"loss": 0.5578,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.15322580933570862,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 752.0,
|
| 615 |
+
"valid_targets_min": 463
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7872,
|
| 619 |
+
"grad_norm": 0.640918224676768,
|
| 620 |
+
"learning_rate": 3.258871313677274e-05,
|
| 621 |
+
"loss": 0.5599,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.13823790848255157,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 789.4,
|
| 626 |
+
"valid_targets_min": 321
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8192,
|
| 630 |
+
"grad_norm": 0.6654266115883691,
|
| 631 |
+
"learning_rate": 3.2239849310778316e-05,
|
| 632 |
+
"loss": 0.5485,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.14035166800022125,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 807.4,
|
| 637 |
+
"valid_targets_min": 448
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.8512,
|
| 641 |
+
"grad_norm": 0.6123014675225724,
|
| 642 |
+
"learning_rate": 3.188492665840909e-05,
|
| 643 |
+
"loss": 0.5557,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.12820711731910706,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 850.4,
|
| 648 |
+
"valid_targets_min": 424
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.8832,
|
| 652 |
+
"grad_norm": 0.6754595695415974,
|
| 653 |
+
"learning_rate": 3.1524120869305726e-05,
|
| 654 |
+
"loss": 0.5627,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.15113355219364166,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 696.0,
|
| 659 |
+
"valid_targets_min": 472
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.9152,
|
| 663 |
+
"grad_norm": 0.6578705510880959,
|
| 664 |
+
"learning_rate": 3.11576105453101e-05,
|
| 665 |
+
"loss": 0.5519,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.15135137736797333,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 767.6,
|
| 670 |
+
"valid_targets_min": 517
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.9472,
|
| 674 |
+
"grad_norm": 0.6449455299555445,
|
| 675 |
+
"learning_rate": 3.0785577112055916e-05,
|
| 676 |
+
"loss": 0.548,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.14119261503219604,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 762.3,
|
| 681 |
+
"valid_targets_min": 434
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9792,
|
| 685 |
+
"grad_norm": 0.6408490242266223,
|
| 686 |
+
"learning_rate": 3.040820472916153e-05,
|
| 687 |
+
"loss": 0.5456,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.12257601320743561,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 696.9,
|
| 692 |
+
"valid_targets_min": 300
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.0064,
|
| 696 |
+
"grad_norm": 0.7541041516650534,
|
| 697 |
+
"learning_rate": 3.002568019906939e-05,
|
| 698 |
+
"loss": 0.5427,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.15626591444015503,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 719.5,
|
| 703 |
+
"valid_targets_min": 396
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0384,
|
| 707 |
+
"grad_norm": 0.665090463756649,
|
| 708 |
+
"learning_rate": 2.963819287457733e-05,
|
| 709 |
+
"loss": 0.5352,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.13162535429000854,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 670.2,
|
| 714 |
+
"valid_targets_min": 440
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0704,
|
| 718 |
+
"grad_norm": 0.6366823920807368,
|
| 719 |
+
"learning_rate": 2.924593456510733e-05,
|
| 720 |
+
"loss": 0.5321,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.13383950293064117,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 810.5,
|
| 725 |
+
"valid_targets_min": 394
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.1024,
|
| 729 |
+
"grad_norm": 0.6213548819315227,
|
| 730 |
+
"learning_rate": 2.8849099441758306e-05,
|
| 731 |
+
"loss": 0.5265,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.10977214574813843,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 728.4,
|
| 736 |
+
"valid_targets_min": 429
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.1344,
|
| 740 |
+
"grad_norm": 0.6547366858654076,
|
| 741 |
+
"learning_rate": 2.844788394118979e-05,
|
| 742 |
+
"loss": 0.5254,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.1424219012260437,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 800.0,
|
| 747 |
+
"valid_targets_min": 452
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.1664,
|
| 751 |
+
"grad_norm": 0.6619785188617626,
|
| 752 |
+
"learning_rate": 2.8042486668384164e-05,
|
| 753 |
+
"loss": 0.5306,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.12859517335891724,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 695.9,
|
| 758 |
+
"valid_targets_min": 435
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.1984,
|
| 762 |
+
"grad_norm": 0.687101029751781,
|
| 763 |
+
"learning_rate": 2.7633108298335582e-05,
|
| 764 |
+
"loss": 0.5184,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.12573498487472534,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 777.2,
|
| 769 |
+
"valid_targets_min": 532
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.2304,
|
| 773 |
+
"grad_norm": 0.645409223244431,
|
| 774 |
+
"learning_rate": 2.721995147671416e-05,
|
| 775 |
+
"loss": 0.5215,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.118968665599823,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 712.1,
|
| 780 |
+
"valid_targets_min": 426
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.2624,
|
| 784 |
+
"grad_norm": 0.671387281197927,
|
| 785 |
+
"learning_rate": 2.68032207195547e-05,
|
| 786 |
+
"loss": 0.5203,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.12927383184432983,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 742.2,
|
| 791 |
+
"valid_targets_min": 464
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.2944,
|
| 795 |
+
"grad_norm": 0.7041426331177664,
|
| 796 |
+
"learning_rate": 2.6383122312019604e-05,
|
| 797 |
+
"loss": 0.5165,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.13757221400737762,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 762.4,
|
| 802 |
+
"valid_targets_min": 397
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3264,
|
| 806 |
+
"grad_norm": 0.5959539114433047,
|
| 807 |
+
"learning_rate": 2.595986420628597e-05,
|
| 808 |
+
"loss": 0.5197,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.1234401986002922,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 834.3,
|
| 813 |
+
"valid_targets_min": 459
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3584,
|
| 817 |
+
"grad_norm": 0.6531185712082775,
|
| 818 |
+
"learning_rate": 2.5533655918607573e-05,
|
| 819 |
+
"loss": 0.5098,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.1260676383972168,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 682.6,
|
| 824 |
+
"valid_targets_min": 395
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.3904,
|
| 828 |
+
"grad_norm": 0.6634208938956202,
|
| 829 |
+
"learning_rate": 2.510470842560259e-05,
|
| 830 |
+
"loss": 0.533,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.1401265263557434,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 758.4,
|
| 835 |
+
"valid_targets_min": 325
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4224,
|
| 839 |
+
"grad_norm": 0.6727520164532284,
|
| 840 |
+
"learning_rate": 2.467323405981841e-05,
|
| 841 |
+
"loss": 0.5271,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.13311098515987396,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 742.8,
|
| 846 |
+
"valid_targets_min": 454
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.4544,
|
| 850 |
+
"grad_norm": 0.6427726949014289,
|
| 851 |
+
"learning_rate": 2.423944640462533e-05,
|
| 852 |
+
"loss": 0.5202,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.13644330203533173,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 789.9,
|
| 857 |
+
"valid_targets_min": 428
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.4864,
|
| 861 |
+
"grad_norm": 0.6840033417316511,
|
| 862 |
+
"learning_rate": 2.3803560188490968e-05,
|
| 863 |
+
"loss": 0.5308,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.12673068046569824,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 702.9,
|
| 868 |
+
"valid_targets_min": 420
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.5183999999999997,
|
| 872 |
+
"grad_norm": 0.6563061162401826,
|
| 873 |
+
"learning_rate": 2.336579117868789e-05,
|
| 874 |
+
"loss": 0.5297,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.1355002373456955,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 729.9,
|
| 879 |
+
"valid_targets_min": 526
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5504,
|
| 883 |
+
"grad_norm": 0.641354939892573,
|
| 884 |
+
"learning_rate": 2.292635607448711e-05,
|
| 885 |
+
"loss": 0.5177,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.12141138315200806,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 733.4,
|
| 890 |
+
"valid_targets_min": 316
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.5824,
|
| 894 |
+
"grad_norm": 0.8906863552430406,
|
| 895 |
+
"learning_rate": 2.248547239989008e-05,
|
| 896 |
+
"loss": 0.5165,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.12363258004188538,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 859.4,
|
| 901 |
+
"valid_targets_min": 384
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.6144,
|
| 905 |
+
"grad_norm": 0.676141962708069,
|
| 906 |
+
"learning_rate": 2.204335839595255e-05,
|
| 907 |
+
"loss": 0.5358,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.14006322622299194,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 766.3,
|
| 912 |
+
"valid_targets_min": 300
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6464,
|
| 916 |
+
"grad_norm": 0.615596066669323,
|
| 917 |
+
"learning_rate": 2.1600232912753452e-05,
|
| 918 |
+
"loss": 0.5215,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.12042121589183807,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 843.1,
|
| 923 |
+
"valid_targets_min": 342
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.6784,
|
| 927 |
+
"grad_norm": 0.615943796645789,
|
| 928 |
+
"learning_rate": 2.1156315301062293e-05,
|
| 929 |
+
"loss": 0.5144,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.1382143646478653,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 864.3,
|
| 934 |
+
"valid_targets_min": 464
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.7104,
|
| 938 |
+
"grad_norm": 1.1772547291108486,
|
| 939 |
+
"learning_rate": 2.0711825303758712e-05,
|
| 940 |
+
"loss": 0.535,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.1458306610584259,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 898.8,
|
| 945 |
+
"valid_targets_min": 554
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7424,
|
| 949 |
+
"grad_norm": 0.6393166323484019,
|
| 950 |
+
"learning_rate": 2.0266982947057962e-05,
|
| 951 |
+
"loss": 0.5235,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.13645440340042114,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 740.2,
|
| 956 |
+
"valid_targets_min": 391
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.7744,
|
| 960 |
+
"grad_norm": 0.6786636785060219,
|
| 961 |
+
"learning_rate": 1.9822008431596083e-05,
|
| 962 |
+
"loss": 0.5337,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.14586326479911804,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 734.4,
|
| 967 |
+
"valid_targets_min": 313
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.8064,
|
| 971 |
+
"grad_norm": 0.6138448927187343,
|
| 972 |
+
"learning_rate": 1.937712202342881e-05,
|
| 973 |
+
"loss": 0.5094,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.14957866072654724,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 983.4,
|
| 978 |
+
"valid_targets_min": 476
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8384,
|
| 982 |
+
"grad_norm": 0.6115317247558727,
|
| 983 |
+
"learning_rate": 1.8932543944998037e-05,
|
| 984 |
+
"loss": 0.5215,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.13156554102897644,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 756.7,
|
| 989 |
+
"valid_targets_min": 509
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.8704,
|
| 993 |
+
"grad_norm": 0.6480026309160363,
|
| 994 |
+
"learning_rate": 1.8488494266119877e-05,
|
| 995 |
+
"loss": 0.5164,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.13125823438167572,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 749.1,
|
| 1000 |
+
"valid_targets_min": 384
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.9024,
|
| 1004 |
+
"grad_norm": 0.6221700405662001,
|
| 1005 |
+
"learning_rate": 1.804519279504834e-05,
|
| 1006 |
+
"loss": 0.5289,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.1366758495569229,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 716.8,
|
| 1011 |
+
"valid_targets_min": 372
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9344,
|
| 1015 |
+
"grad_norm": 0.6835135386375543,
|
| 1016 |
+
"learning_rate": 1.7602858969668365e-05,
|
| 1017 |
+
"loss": 0.5423,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.1429317593574524,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 738.8,
|
| 1022 |
+
"valid_targets_min": 370
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.9664,
|
| 1026 |
+
"grad_norm": 0.6476761584366846,
|
| 1027 |
+
"learning_rate": 1.716171174887231e-05,
|
| 1028 |
+
"loss": 0.5177,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.12608763575553894,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 771.1,
|
| 1033 |
+
"valid_targets_min": 577
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.9984,
|
| 1037 |
+
"grad_norm": 0.6705239133879689,
|
| 1038 |
+
"learning_rate": 1.6721969504173484e-05,
|
| 1039 |
+
"loss": 0.5304,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.156068354845047,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 891.4,
|
| 1044 |
+
"valid_targets_min": 436
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0256,
|
| 1048 |
+
"grad_norm": 0.6254648310969626,
|
| 1049 |
+
"learning_rate": 1.628384991161041e-05,
|
| 1050 |
+
"loss": 0.4947,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.12908436357975006,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 856.1,
|
| 1055 |
+
"valid_targets_min": 468
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.0576,
|
| 1059 |
+
"grad_norm": 0.6835751653097379,
|
| 1060 |
+
"learning_rate": 1.5847569843995452e-05,
|
| 1061 |
+
"loss": 0.5142,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.14210891723632812,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 753.6,
|
| 1066 |
+
"valid_targets_min": 416
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.0896,
|
| 1070 |
+
"grad_norm": 0.6619131317365188,
|
| 1071 |
+
"learning_rate": 1.5413345263560922e-05,
|
| 1072 |
+
"loss": 0.5013,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.12965424358844757,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 726.4,
|
| 1077 |
+
"valid_targets_min": 444
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.1216,
|
| 1081 |
+
"grad_norm": 0.6591313448070079,
|
| 1082 |
+
"learning_rate": 1.4981391115056032e-05,
|
| 1083 |
+
"loss": 0.5063,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.1259090006351471,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 749.0,
|
| 1088 |
+
"valid_targets_min": 385
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1536,
|
| 1092 |
+
"grad_norm": 0.6569116597796409,
|
| 1093 |
+
"learning_rate": 1.455192121934748e-05,
|
| 1094 |
+
"loss": 0.5012,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.1285029649734497,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 837.4,
|
| 1099 |
+
"valid_targets_min": 496
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.1856,
|
| 1103 |
+
"grad_norm": 0.6316910437995757,
|
| 1104 |
+
"learning_rate": 1.4125148167576303e-05,
|
| 1105 |
+
"loss": 0.5031,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.11511541903018951,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 672.5,
|
| 1110 |
+
"valid_targets_min": 430
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.2176,
|
| 1114 |
+
"grad_norm": 0.6389505729458541,
|
| 1115 |
+
"learning_rate": 1.3701283215923563e-05,
|
| 1116 |
+
"loss": 0.5044,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.12613186240196228,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 796.8,
|
| 1121 |
+
"valid_targets_min": 310
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2496,
|
| 1125 |
+
"grad_norm": 0.6629145801717702,
|
| 1126 |
+
"learning_rate": 1.328053618103677e-05,
|
| 1127 |
+
"loss": 0.495,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.11542559415102005,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 727.7,
|
| 1132 |
+
"valid_targets_min": 439
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.2816,
|
| 1136 |
+
"grad_norm": 0.634372579989713,
|
| 1137 |
+
"learning_rate": 1.2863115336168916e-05,
|
| 1138 |
+
"loss": 0.5014,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.12187671661376953,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 753.3,
|
| 1143 |
+
"valid_targets_min": 482
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3136,
|
| 1147 |
+
"grad_norm": 0.6384070283786387,
|
| 1148 |
+
"learning_rate": 1.2449227308081509e-05,
|
| 1149 |
+
"loss": 0.5008,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.1421659290790558,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 845.1,
|
| 1154 |
+
"valid_targets_min": 573
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.3456,
|
| 1158 |
+
"grad_norm": 0.6332388154728119,
|
| 1159 |
+
"learning_rate": 1.2039076974762587e-05,
|
| 1160 |
+
"loss": 0.4928,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.11383464932441711,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 718.1,
|
| 1165 |
+
"valid_targets_min": 397
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.3776,
|
| 1169 |
+
"grad_norm": 0.6093799616673348,
|
| 1170 |
+
"learning_rate": 1.163286736401044e-05,
|
| 1171 |
+
"loss": 0.4926,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.12254788726568222,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 798.4,
|
| 1176 |
+
"valid_targets_min": 413
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4096,
|
| 1180 |
+
"grad_norm": 0.6455046460858423,
|
| 1181 |
+
"learning_rate": 1.123079955293322e-05,
|
| 1182 |
+
"loss": 0.5112,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.13581742346286774,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 846.3,
|
| 1187 |
+
"valid_targets_min": 520
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.4416,
|
| 1191 |
+
"grad_norm": 0.631427790063437,
|
| 1192 |
+
"learning_rate": 1.0833072568414037e-05,
|
| 1193 |
+
"loss": 0.5119,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.1291504204273224,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 806.6,
|
| 1198 |
+
"valid_targets_min": 501
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.4736000000000002,
|
| 1202 |
+
"grad_norm": 0.6322674040782007,
|
| 1203 |
+
"learning_rate": 1.0439883288591057e-05,
|
| 1204 |
+
"loss": 0.4964,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.11973114311695099,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 736.1,
|
| 1209 |
+
"valid_targets_min": 400
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.5056000000000003,
|
| 1213 |
+
"grad_norm": 0.6441453581221164,
|
| 1214 |
+
"learning_rate": 1.0051426345401202e-05,
|
| 1215 |
+
"loss": 0.5132,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.13486211001873016,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 804.2,
|
| 1220 |
+
"valid_targets_min": 478
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5376,
|
| 1224 |
+
"grad_norm": 0.6425950808621582,
|
| 1225 |
+
"learning_rate": 9.667894028235704e-06,
|
| 1226 |
+
"loss": 0.5077,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.12831318378448486,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 724.8,
|
| 1231 |
+
"valid_targets_min": 512
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.5696,
|
| 1235 |
+
"grad_norm": 0.6124823375637398,
|
| 1236 |
+
"learning_rate": 9.289476188755315e-06,
|
| 1237 |
+
"loss": 0.5038,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.12071146070957184,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 741.8,
|
| 1242 |
+
"valid_targets_min": 316
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6016,
|
| 1246 |
+
"grad_norm": 0.6205786442665068,
|
| 1247 |
+
"learning_rate": 8.916360146912122e-06,
|
| 1248 |
+
"loss": 0.5037,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.10591184347867966,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 709.2,
|
| 1253 |
+
"valid_targets_min": 480
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.6336,
|
| 1257 |
+
"grad_norm": 0.6636478895992294,
|
| 1258 |
+
"learning_rate": 8.548730598224646e-06,
|
| 1259 |
+
"loss": 0.4993,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.12058466672897339,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 727.6,
|
| 1264 |
+
"valid_targets_min": 417
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.6656,
|
| 1268 |
+
"grad_norm": 0.5971242481772463,
|
| 1269 |
+
"learning_rate": 8.186769522352053e-06,
|
| 1270 |
+
"loss": 0.4952,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.13874907791614532,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 860.1,
|
| 1275 |
+
"valid_targets_min": 527
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.6976,
|
| 1279 |
+
"grad_norm": 0.615075544398069,
|
| 1280 |
+
"learning_rate": 7.830656093012714e-06,
|
| 1281 |
+
"loss": 0.5037,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.11050853878259659,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 743.3,
|
| 1286 |
+
"valid_targets_min": 468
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.7296,
|
| 1290 |
+
"grad_norm": 0.6470427339088323,
|
| 1291 |
+
"learning_rate": 7.480566589291696e-06,
|
| 1292 |
+
"loss": 0.5064,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.13953588902950287,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 822.9,
|
| 1297 |
+
"valid_targets_min": 493
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.7616,
|
| 1301 |
+
"grad_norm": 0.597610350393088,
|
| 1302 |
+
"learning_rate": 7.1366743083812285e-06,
|
| 1303 |
+
"loss": 0.4967,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.11674021184444427,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 738.9,
|
| 1308 |
+
"valid_targets_min": 415
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.7936,
|
| 1312 |
+
"grad_norm": 0.6338991358688321,
|
| 1313 |
+
"learning_rate": 6.799149479797101e-06,
|
| 1314 |
+
"loss": 0.5006,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.12019048631191254,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 707.8,
|
| 1319 |
+
"valid_targets_min": 442
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.8256,
|
| 1323 |
+
"grad_norm": 0.6202165939689909,
|
| 1324 |
+
"learning_rate": 6.4681591811137e-06,
|
| 1325 |
+
"loss": 0.4968,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.12805257737636566,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 873.7,
|
| 1330 |
+
"valid_targets_min": 529
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.8576,
|
| 1334 |
+
"grad_norm": 0.6714289302846277,
|
| 1335 |
+
"learning_rate": 6.143867255259197e-06,
|
| 1336 |
+
"loss": 0.5012,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.12346205115318298,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 609.8,
|
| 1341 |
+
"valid_targets_min": 416
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.8895999999999997,
|
| 1345 |
+
"grad_norm": 0.6242136384257845,
|
| 1346 |
+
"learning_rate": 5.8264342294119504e-06,
|
| 1347 |
+
"loss": 0.501,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.11968934535980225,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 689.3,
|
| 1352 |
+
"valid_targets_min": 443
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.9215999999999998,
|
| 1356 |
+
"grad_norm": 0.6463661477202667,
|
| 1357 |
+
"learning_rate": 5.516017235538258e-06,
|
| 1358 |
+
"loss": 0.5053,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.1359320878982544,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 786.9,
|
| 1363 |
+
"valid_targets_min": 559
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.9536,
|
| 1367 |
+
"grad_norm": 0.6491908587833649,
|
| 1368 |
+
"learning_rate": 5.212769932610695e-06,
|
| 1369 |
+
"loss": 0.4973,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.14212754368782043,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 861.1,
|
| 1374 |
+
"valid_targets_min": 507
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.9856,
|
| 1378 |
+
"grad_norm": 0.6281434470761624,
|
| 1379 |
+
"learning_rate": 4.916842430545681e-06,
|
| 1380 |
+
"loss": 0.4975,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.11817573010921478,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 755.6,
|
| 1385 |
+
"valid_targets_min": 445
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.0128,
|
| 1389 |
+
"grad_norm": 0.6490532684460447,
|
| 1390 |
+
"learning_rate": 4.628381215897837e-06,
|
| 1391 |
+
"loss": 0.4819,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.11554375290870667,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 724.1,
|
| 1396 |
+
"valid_targets_min": 367
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.0448,
|
| 1400 |
+
"grad_norm": 0.6327445790214326,
|
| 1401 |
+
"learning_rate": 4.347529079347914e-06,
|
| 1402 |
+
"loss": 0.4894,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.12214243412017822,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 676.2,
|
| 1407 |
+
"valid_targets_min": 310
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.0768,
|
| 1411 |
+
"grad_norm": 0.6158265147921352,
|
| 1412 |
+
"learning_rate": 4.074425045020247e-06,
|
| 1413 |
+
"loss": 0.4868,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.13750189542770386,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 834.1,
|
| 1418 |
+
"valid_targets_min": 555
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.1088,
|
| 1422 |
+
"grad_norm": 0.627070881496122,
|
| 1423 |
+
"learning_rate": 3.8092043016646487e-06,
|
| 1424 |
+
"loss": 0.4929,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.11655047535896301,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 674.8,
|
| 1429 |
+
"valid_targets_min": 411
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.1408,
|
| 1433 |
+
"grad_norm": 0.6221973742999799,
|
| 1434 |
+
"learning_rate": 3.551998135736867e-06,
|
| 1435 |
+
"loss": 0.489,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.12250182032585144,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 665.8,
|
| 1440 |
+
"valid_targets_min": 394
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.1728,
|
| 1444 |
+
"grad_norm": 0.6626884060775692,
|
| 1445 |
+
"learning_rate": 3.3029338664107267e-06,
|
| 1446 |
+
"loss": 0.4741,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.12655746936798096,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 705.9,
|
| 1451 |
+
"valid_targets_min": 548
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.2048,
|
| 1455 |
+
"grad_norm": 0.6224489306449362,
|
| 1456 |
+
"learning_rate": 3.0621347825540625e-06,
|
| 1457 |
+
"loss": 0.4841,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.12335214763879776,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 784.9,
|
| 1462 |
+
"valid_targets_min": 525
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.2368,
|
| 1466 |
+
"grad_norm": 0.6520842934256829,
|
| 1467 |
+
"learning_rate": 2.8297200816997183e-06,
|
| 1468 |
+
"loss": 0.4911,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.12841469049453735,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 765.9,
|
| 1473 |
+
"valid_targets_min": 481
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.2688,
|
| 1477 |
+
"grad_norm": 0.6423816689951226,
|
| 1478 |
+
"learning_rate": 2.605804811041803e-06,
|
| 1479 |
+
"loss": 0.4819,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.10488501191139221,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 732.0,
|
| 1484 |
+
"valid_targets_min": 450
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.3008,
|
| 1488 |
+
"grad_norm": 0.6181615169830669,
|
| 1489 |
+
"learning_rate": 2.390499810486351e-06,
|
| 1490 |
+
"loss": 0.481,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.12980340421199799,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 780.8,
|
| 1495 |
+
"valid_targets_min": 466
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.3328,
|
| 1499 |
+
"grad_norm": 0.6322677157647695,
|
| 1500 |
+
"learning_rate": 2.183911657784685e-06,
|
| 1501 |
+
"loss": 0.4964,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.12320327758789062,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 868.0,
|
| 1506 |
+
"valid_targets_min": 429
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.3648,
|
| 1510 |
+
"grad_norm": 0.6043179861396448,
|
| 1511 |
+
"learning_rate": 1.986142615776532e-06,
|
| 1512 |
+
"loss": 0.4919,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.1195220947265625,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 728.2,
|
| 1517 |
+
"valid_targets_min": 471
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.3968,
|
| 1521 |
+
"grad_norm": 0.640478901066271,
|
| 1522 |
+
"learning_rate": 1.7972905817690644e-06,
|
| 1523 |
+
"loss": 0.4888,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.12453019618988037,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 716.9,
|
| 1528 |
+
"valid_targets_min": 362
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.4288,
|
| 1532 |
+
"grad_norm": 0.6112657891179784,
|
| 1533 |
+
"learning_rate": 1.617449039076955e-06,
|
| 1534 |
+
"loss": 0.4972,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.12277944386005402,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 826.2,
|
| 1539 |
+
"valid_targets_min": 418
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.4608,
|
| 1543 |
+
"grad_norm": 0.680661377919841,
|
| 1544 |
+
"learning_rate": 1.4467070107473413e-06,
|
| 1545 |
+
"loss": 0.4926,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.11709259450435638,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 670.3,
|
| 1550 |
+
"valid_targets_min": 446
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.4928,
|
| 1554 |
+
"grad_norm": 0.6067946877348196,
|
| 1555 |
+
"learning_rate": 1.2851490154926816e-06,
|
| 1556 |
+
"loss": 0.4844,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.11835283041000366,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 833.8,
|
| 1561 |
+
"valid_targets_min": 535
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.5248,
|
| 1565 |
+
"grad_norm": 0.6214656228794159,
|
| 1566 |
+
"learning_rate": 1.1328550258533211e-06,
|
| 1567 |
+
"loss": 0.4955,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.1317848563194275,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 780.4,
|
| 1572 |
+
"valid_targets_min": 540
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.5568,
|
| 1576 |
+
"grad_norm": 0.6101425589821468,
|
| 1577 |
+
"learning_rate": 9.899004286103953e-07,
|
| 1578 |
+
"loss": 0.4892,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.1342540979385376,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 813.8,
|
| 1583 |
+
"valid_targets_min": 592
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.5888,
|
| 1587 |
+
"grad_norm": 0.6431480087423993,
|
| 1588 |
+
"learning_rate": 8.5635598746876e-07,
|
| 1589 |
+
"loss": 0.4892,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.12438573688268661,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 771.0,
|
| 1594 |
+
"valid_targets_min": 505
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.6208,
|
| 1598 |
+
"grad_norm": 0.6403336841979095,
|
| 1599 |
+
"learning_rate": 7.32287808028389e-07,
|
| 1600 |
+
"loss": 0.502,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.14776507019996643,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 895.8,
|
| 1605 |
+
"valid_targets_min": 484
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.6528,
|
| 1609 |
+
"grad_norm": 0.6038822017761685,
|
| 1610 |
+
"learning_rate": 6.177573050615327e-07,
|
| 1611 |
+
"loss": 0.4909,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.11538829654455185,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 747.9,
|
| 1616 |
+
"valid_targets_min": 428
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.6848,
|
| 1620 |
+
"grad_norm": 0.6324119113515146,
|
| 1621 |
+
"learning_rate": 5.128211721119213e-07,
|
| 1622 |
+
"loss": 0.4979,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.13889926671981812,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 792.2,
|
| 1627 |
+
"valid_targets_min": 385
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.7168,
|
| 1631 |
+
"grad_norm": 0.6324936257203374,
|
| 1632 |
+
"learning_rate": 4.175313534309755e-07,
|
| 1633 |
+
"loss": 0.487,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.1103043407201767,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 703.1,
|
| 1638 |
+
"valid_targets_min": 435
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.7488,
|
| 1642 |
+
"grad_norm": 0.629729044732728,
|
| 1643 |
+
"learning_rate": 3.319350182649861e-07,
|
| 1644 |
+
"loss": 0.4892,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.12301481515169144,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 739.1,
|
| 1649 |
+
"valid_targets_min": 491
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.7808,
|
| 1653 |
+
"grad_norm": 0.6274760958790774,
|
| 1654 |
+
"learning_rate": 2.560745375059392e-07,
|
| 1655 |
+
"loss": 0.4877,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.11202463507652283,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 644.5,
|
| 1660 |
+
"valid_targets_min": 356
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.8128,
|
| 1664 |
+
"grad_norm": 0.6073341940792345,
|
| 1665 |
+
"learning_rate": 1.8998746271758016e-07,
|
| 1666 |
+
"loss": 0.4851,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.12768879532814026,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 871.2,
|
| 1671 |
+
"valid_targets_min": 439
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.8448,
|
| 1675 |
+
"grad_norm": 0.6300616469590965,
|
| 1676 |
+
"learning_rate": 1.337065075470778e-07,
|
| 1677 |
+
"loss": 0.4962,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.10540274530649185,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 732.8,
|
| 1682 |
+
"valid_targets_min": 458
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.8768,
|
| 1686 |
+
"grad_norm": 0.607465565387994,
|
| 1687 |
+
"learning_rate": 8.725953153150279e-08,
|
| 1688 |
+
"loss": 0.4911,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.132490336894989,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 854.8,
|
| 1693 |
+
"valid_targets_min": 435
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.9088,
|
| 1697 |
+
"grad_norm": 0.5917313437808399,
|
| 1698 |
+
"learning_rate": 5.066952630711886e-08,
|
| 1699 |
+
"loss": 0.4864,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.10872993618249893,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 783.2,
|
| 1704 |
+
"valid_targets_min": 507
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.9408,
|
| 1708 |
+
"grad_norm": 0.7403546016233938,
|
| 1709 |
+
"learning_rate": 2.3954604228342283e-08,
|
| 1710 |
+
"loss": 0.4979,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.12833866477012634,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 677.8,
|
| 1715 |
+
"valid_targets_min": 428
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.9728,
|
| 1719 |
+
"grad_norm": 0.6407918253430359,
|
| 1720 |
+
"learning_rate": 7.12798940197601e-09,
|
| 1721 |
+
"loss": 0.498,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.12702913582324982,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 762.4,
|
| 1726 |
+
"valid_targets_min": 411
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"grad_norm": 1.2926809123761072,
|
| 1731 |
+
"learning_rate": 1.9801114115480802e-10,
|
| 1732 |
+
"loss": 0.4833,
|
| 1733 |
+
"loss_nan_ranks": 0,
|
| 1734 |
+
"loss_rank_avg": 0.4620312452316284,
|
| 1735 |
+
"step": 785,
|
| 1736 |
+
"valid_targets_mean": 702.6,
|
| 1737 |
+
"valid_targets_min": 459
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 5.0,
|
| 1741 |
+
"loss_nan_ranks": 0,
|
| 1742 |
+
"loss_rank_avg": 0.4620312452316284,
|
| 1743 |
+
"step": 785,
|
| 1744 |
+
"total_flos": 2.3852276139845222e+17,
|
| 1745 |
+
"train_loss": 0.5620896111628052,
|
| 1746 |
+
"train_runtime": 6728.981,
|
| 1747 |
+
"train_samples_per_second": 7.428,
|
| 1748 |
+
"train_steps_per_second": 0.117,
|
| 1749 |
+
"valid_targets_mean": 702.6,
|
| 1750 |
+
"valid_targets_min": 459
|
| 1751 |
+
}
|
| 1752 |
+
],
|
| 1753 |
+
"logging_steps": 5,
|
| 1754 |
+
"max_steps": 785,
|
| 1755 |
+
"num_input_tokens_seen": 0,
|
| 1756 |
+
"num_train_epochs": 5,
|
| 1757 |
+
"save_steps": 500,
|
| 1758 |
+
"stateful_callbacks": {
|
| 1759 |
+
"TrainerControl": {
|
| 1760 |
+
"args": {
|
| 1761 |
+
"should_epoch_stop": false,
|
| 1762 |
+
"should_evaluate": false,
|
| 1763 |
+
"should_log": false,
|
| 1764 |
+
"should_save": false,
|
| 1765 |
+
"should_training_stop": false
|
| 1766 |
+
},
|
| 1767 |
+
"attributes": {}
|
| 1768 |
+
}
|
| 1769 |
+
},
|
| 1770 |
+
"total_flos": 2.3852276139845222e+17,
|
| 1771 |
+
"train_batch_size": 1,
|
| 1772 |
+
"trial_name": null,
|
| 1773 |
+
"trial_params": null
|
| 1774 |
+
}
|
training_loss.png
ADDED
|