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-sandboxes-traces-terminus-2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# taskmaster2-sandboxes-traces-terminus-2
|
| 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-sandboxes-traces-terminus-2
|
|
|
|
| 16 |
|
| 17 |
# taskmaster2-sandboxes-traces-terminus-2
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the mlfoundations-dev/taskmaster2-sandboxes-traces-terminus-2 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.805999622358391,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 152.45570532704255,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.40260475873947144,
|
| 7 |
+
"mfu_percent": 0.33964661642108773,
|
| 8 |
+
"mfu_percent_theoretical": 10.77425479343057,
|
| 9 |
+
"total_flos": 1.3461651925678162e+18,
|
| 10 |
+
"train_loss": 0.41955339954157544,
|
| 11 |
+
"train_runtime": 17506.3111,
|
| 12 |
+
"train_samples_per_second": 2.856,
|
| 13 |
+
"train_steps_per_second": 0.045,
|
| 14 |
+
"valid_targets_mean": 3870.8,
|
| 15 |
+
"valid_targets_min": 1612
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.805999622358391,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 152.45570532704255,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.40260475873947144,
|
| 7 |
+
"mfu_percent": 0.33964661642108773,
|
| 8 |
+
"mfu_percent_theoretical": 10.77425479343057,
|
| 9 |
+
"total_flos": 1.3461651925678162e+18,
|
| 10 |
+
"train_loss": 0.41955339954157544,
|
| 11 |
+
"train_runtime": 17506.3111,
|
| 12 |
+
"train_samples_per_second": 2.856,
|
| 13 |
+
"train_steps_per_second": 0.045,
|
| 14 |
+
"valid_targets_mean": 3870.8,
|
| 15 |
+
"valid_targets_min": 1612
|
| 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": 6.048756324166434,
|
| 15 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 16 |
+
"loss": 0.7749,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.1723039150238037,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 3899.1,
|
| 21 |
+
"valid_targets_min": 639
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.064,
|
| 25 |
+
"grad_norm": 3.820317754564466,
|
| 26 |
+
"learning_rate": 4.556962025316456e-06,
|
| 27 |
+
"loss": 0.7446,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.185373917222023,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 5002.5,
|
| 32 |
+
"valid_targets_min": 1422
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.096,
|
| 36 |
+
"grad_norm": 2.5037828274167304,
|
| 37 |
+
"learning_rate": 7.08860759493671e-06,
|
| 38 |
+
"loss": 0.6931,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.17158135771751404,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 3496.6,
|
| 43 |
+
"valid_targets_min": 1215
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.128,
|
| 47 |
+
"grad_norm": 0.9315157784215937,
|
| 48 |
+
"learning_rate": 9.620253164556963e-06,
|
| 49 |
+
"loss": 0.6545,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.17365200817584991,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 4261.8,
|
| 54 |
+
"valid_targets_min": 1995
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.16,
|
| 58 |
+
"grad_norm": 0.7754812519149545,
|
| 59 |
+
"learning_rate": 1.2151898734177216e-05,
|
| 60 |
+
"loss": 0.611,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.12669777870178223,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 2931.9,
|
| 65 |
+
"valid_targets_min": 566
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.192,
|
| 69 |
+
"grad_norm": 0.52185829607096,
|
| 70 |
+
"learning_rate": 1.468354430379747e-05,
|
| 71 |
+
"loss": 0.5995,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.14718282222747803,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 4264.7,
|
| 76 |
+
"valid_targets_min": 1204
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.224,
|
| 80 |
+
"grad_norm": 0.4852281885120764,
|
| 81 |
+
"learning_rate": 1.7215189873417723e-05,
|
| 82 |
+
"loss": 0.5804,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.13390770554542542,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 3495.6,
|
| 87 |
+
"valid_targets_min": 600
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.256,
|
| 91 |
+
"grad_norm": 0.38657633820848264,
|
| 92 |
+
"learning_rate": 1.974683544303798e-05,
|
| 93 |
+
"loss": 0.5536,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.11915287375450134,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 4156.8,
|
| 98 |
+
"valid_targets_min": 816
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.288,
|
| 102 |
+
"grad_norm": 0.3612407081777249,
|
| 103 |
+
"learning_rate": 2.2278481012658228e-05,
|
| 104 |
+
"loss": 0.5628,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.15402325987815857,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 4639.7,
|
| 109 |
+
"valid_targets_min": 982
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32,
|
| 113 |
+
"grad_norm": 0.32318118714364474,
|
| 114 |
+
"learning_rate": 2.481012658227848e-05,
|
| 115 |
+
"loss": 0.522,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.13298863172531128,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 3814.6,
|
| 120 |
+
"valid_targets_min": 655
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.352,
|
| 124 |
+
"grad_norm": 0.29866719779423245,
|
| 125 |
+
"learning_rate": 2.7341772151898737e-05,
|
| 126 |
+
"loss": 0.4995,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.12222938239574432,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 3945.7,
|
| 131 |
+
"valid_targets_min": 791
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.384,
|
| 135 |
+
"grad_norm": 0.30726596921009003,
|
| 136 |
+
"learning_rate": 2.987341772151899e-05,
|
| 137 |
+
"loss": 0.4916,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.11845584958791733,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 3824.1,
|
| 142 |
+
"valid_targets_min": 755
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.416,
|
| 146 |
+
"grad_norm": 0.2521297207306386,
|
| 147 |
+
"learning_rate": 3.240506329113924e-05,
|
| 148 |
+
"loss": 0.4989,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.1335579752922058,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 4825.6,
|
| 153 |
+
"valid_targets_min": 2415
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.448,
|
| 157 |
+
"grad_norm": 0.26719933666277185,
|
| 158 |
+
"learning_rate": 3.49367088607595e-05,
|
| 159 |
+
"loss": 0.4844,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.12163548916578293,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 4284.9,
|
| 164 |
+
"valid_targets_min": 535
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48,
|
| 168 |
+
"grad_norm": 0.2979141750380719,
|
| 169 |
+
"learning_rate": 3.746835443037975e-05,
|
| 170 |
+
"loss": 0.4583,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.11367259174585342,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 4516.4,
|
| 175 |
+
"valid_targets_min": 911
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.512,
|
| 179 |
+
"grad_norm": 0.30292473042949297,
|
| 180 |
+
"learning_rate": 4e-05,
|
| 181 |
+
"loss": 0.4716,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.12441614270210266,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 4018.8,
|
| 186 |
+
"valid_targets_min": 707
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.544,
|
| 190 |
+
"grad_norm": 0.28266530877838175,
|
| 191 |
+
"learning_rate": 3.999504991751045e-05,
|
| 192 |
+
"loss": 0.4824,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.10960252583026886,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 3031.5,
|
| 197 |
+
"valid_targets_min": 902
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.576,
|
| 201 |
+
"grad_norm": 0.287898903894681,
|
| 202 |
+
"learning_rate": 3.9980202120373464e-05,
|
| 203 |
+
"loss": 0.4586,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.12198155373334885,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 4651.1,
|
| 208 |
+
"valid_targets_min": 1158
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.608,
|
| 212 |
+
"grad_norm": 0.2696418518664897,
|
| 213 |
+
"learning_rate": 3.995546395837111e-05,
|
| 214 |
+
"loss": 0.4502,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.1148751974105835,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 4679.6,
|
| 219 |
+
"valid_targets_min": 485
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.64,
|
| 223 |
+
"grad_norm": 0.29505853573744345,
|
| 224 |
+
"learning_rate": 3.992084767709763e-05,
|
| 225 |
+
"loss": 0.4575,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.10742583870887756,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 3360.1,
|
| 230 |
+
"valid_targets_min": 697
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.672,
|
| 234 |
+
"grad_norm": 0.2941887926453614,
|
| 235 |
+
"learning_rate": 3.987637041189781e-05,
|
| 236 |
+
"loss": 0.4541,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.10949065536260605,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 4291.4,
|
| 241 |
+
"valid_targets_min": 853
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.704,
|
| 245 |
+
"grad_norm": 0.30772503562874043,
|
| 246 |
+
"learning_rate": 3.982205417938482e-05,
|
| 247 |
+
"loss": 0.4564,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.11388377845287323,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 3922.4,
|
| 252 |
+
"valid_targets_min": 586
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.736,
|
| 256 |
+
"grad_norm": 0.2894203299726909,
|
| 257 |
+
"learning_rate": 3.975792586654179e-05,
|
| 258 |
+
"loss": 0.441,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.1191316694021225,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 5007.5,
|
| 263 |
+
"valid_targets_min": 895
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.768,
|
| 267 |
+
"grad_norm": 0.3360267509201506,
|
| 268 |
+
"learning_rate": 3.968401721741259e-05,
|
| 269 |
+
"loss": 0.4523,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.11252386122941971,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 3807.6,
|
| 274 |
+
"valid_targets_min": 783
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8,
|
| 278 |
+
"grad_norm": 0.2954655902404845,
|
| 279 |
+
"learning_rate": 3.960036481738819e-05,
|
| 280 |
+
"loss": 0.4562,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.10796886682510376,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 3645.1,
|
| 285 |
+
"valid_targets_min": 1120
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.832,
|
| 289 |
+
"grad_norm": 0.2979238438193785,
|
| 290 |
+
"learning_rate": 3.950701007509667e-05,
|
| 291 |
+
"loss": 0.4303,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.1278803050518036,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 4754.6,
|
| 296 |
+
"valid_targets_min": 826
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.864,
|
| 300 |
+
"grad_norm": 0.3025824931420387,
|
| 301 |
+
"learning_rate": 3.940399920190552e-05,
|
| 302 |
+
"loss": 0.4324,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.11299727112054825,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 3936.9,
|
| 307 |
+
"valid_targets_min": 651
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.896,
|
| 311 |
+
"grad_norm": 0.27361938948965725,
|
| 312 |
+
"learning_rate": 3.92913831890467e-05,
|
| 313 |
+
"loss": 0.4393,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.11303415894508362,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 4267.7,
|
| 318 |
+
"valid_targets_min": 674
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.928,
|
| 322 |
+
"grad_norm": 0.26468614400279594,
|
| 323 |
+
"learning_rate": 3.916921778237556e-05,
|
| 324 |
+
"loss": 0.4197,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.1154870092868805,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 4747.6,
|
| 329 |
+
"valid_targets_min": 1718
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.96,
|
| 333 |
+
"grad_norm": 0.4577347781033962,
|
| 334 |
+
"learning_rate": 3.903756345477612e-05,
|
| 335 |
+
"loss": 0.4496,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.12438996136188507,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 4225.4,
|
| 340 |
+
"valid_targets_min": 906
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.992,
|
| 344 |
+
"grad_norm": 0.28366351389299993,
|
| 345 |
+
"learning_rate": 3.889648537622657e-05,
|
| 346 |
+
"loss": 0.4352,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.10910991579294205,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 4543.1,
|
| 351 |
+
"valid_targets_min": 723
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0192,
|
| 355 |
+
"grad_norm": 0.2851059883525269,
|
| 356 |
+
"learning_rate": 3.874605338153952e-05,
|
| 357 |
+
"loss": 0.4226,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.09767496585845947,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 3813.6,
|
| 362 |
+
"valid_targets_min": 739
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0512,
|
| 366 |
+
"grad_norm": 0.27633015571631864,
|
| 367 |
+
"learning_rate": 3.8586341935793265e-05,
|
| 368 |
+
"loss": 0.4195,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.09590106457471848,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 4107.2,
|
| 373 |
+
"valid_targets_min": 851
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0832,
|
| 377 |
+
"grad_norm": 0.29862294491293667,
|
| 378 |
+
"learning_rate": 3.841743009747089e-05,
|
| 379 |
+
"loss": 0.4365,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.11445736140012741,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 4095.9,
|
| 384 |
+
"valid_targets_min": 1214
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1152,
|
| 388 |
+
"grad_norm": 0.2665048493087847,
|
| 389 |
+
"learning_rate": 3.8239401479325714e-05,
|
| 390 |
+
"loss": 0.4221,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.10399468243122101,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 4598.8,
|
| 395 |
+
"valid_targets_min": 1231
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1472,
|
| 399 |
+
"grad_norm": 0.3020735192506363,
|
| 400 |
+
"learning_rate": 3.8052344206992276e-05,
|
| 401 |
+
"loss": 0.413,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.11090105026960373,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 4057.2,
|
| 406 |
+
"valid_targets_min": 887
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1792,
|
| 410 |
+
"grad_norm": 0.27625874894849545,
|
| 411 |
+
"learning_rate": 3.7856350875363396e-05,
|
| 412 |
+
"loss": 0.4155,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.09643165022134781,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 4063.6,
|
| 417 |
+
"valid_targets_min": 988
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2112,
|
| 421 |
+
"grad_norm": 0.3022623050650015,
|
| 422 |
+
"learning_rate": 3.765151850275497e-05,
|
| 423 |
+
"loss": 0.4387,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.11312417685985565,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 3933.4,
|
| 428 |
+
"valid_targets_min": 877
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.2432,
|
| 432 |
+
"grad_norm": 0.3109759147204714,
|
| 433 |
+
"learning_rate": 3.7437948482881104e-05,
|
| 434 |
+
"loss": 0.4174,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.11555179208517075,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 3763.6,
|
| 439 |
+
"valid_targets_min": 995
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2752,
|
| 443 |
+
"grad_norm": 0.2923130244414311,
|
| 444 |
+
"learning_rate": 3.721574653466336e-05,
|
| 445 |
+
"loss": 0.4229,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.11472899466753006,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 4125.8,
|
| 450 |
+
"valid_targets_min": 938
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3072,
|
| 454 |
+
"grad_norm": 0.2782752287912167,
|
| 455 |
+
"learning_rate": 3.698502264989903e-05,
|
| 456 |
+
"loss": 0.4211,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.10930650681257248,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 4209.9,
|
| 461 |
+
"valid_targets_min": 542
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.3392,
|
| 465 |
+
"grad_norm": 0.2991828057882393,
|
| 466 |
+
"learning_rate": 3.674589103881432e-05,
|
| 467 |
+
"loss": 0.416,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.10878201574087143,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 4023.2,
|
| 472 |
+
"valid_targets_min": 773
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3712,
|
| 476 |
+
"grad_norm": 0.30997637082473456,
|
| 477 |
+
"learning_rate": 3.64984700735293e-05,
|
| 478 |
+
"loss": 0.4162,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.08643242716789246,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 3282.3,
|
| 483 |
+
"valid_targets_min": 761
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4032,
|
| 487 |
+
"grad_norm": 0.2715643458114872,
|
| 488 |
+
"learning_rate": 3.624288222946273e-05,
|
| 489 |
+
"loss": 0.4217,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.10085868835449219,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 3737.1,
|
| 494 |
+
"valid_targets_min": 661
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4352,
|
| 498 |
+
"grad_norm": 0.3420125787516997,
|
| 499 |
+
"learning_rate": 3.597925402470578e-05,
|
| 500 |
+
"loss": 0.4234,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.11982588469982147,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 4721.4,
|
| 505 |
+
"valid_targets_min": 559
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4672,
|
| 509 |
+
"grad_norm": 0.3500123406955092,
|
| 510 |
+
"learning_rate": 3.570771595739445e-05,
|
| 511 |
+
"loss": 0.4166,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.09180434793233871,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 4225.7,
|
| 516 |
+
"valid_targets_min": 640
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.4992,
|
| 520 |
+
"grad_norm": 0.27575481164813326,
|
| 521 |
+
"learning_rate": 3.5428402441111964e-05,
|
| 522 |
+
"loss": 0.4274,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.09477457404136658,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 3827.4,
|
| 527 |
+
"valid_targets_min": 579
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5312000000000001,
|
| 531 |
+
"grad_norm": 0.308272520000558,
|
| 532 |
+
"learning_rate": 3.5141451738352936e-05,
|
| 533 |
+
"loss": 0.4098,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.1010490283370018,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 4072.6,
|
| 538 |
+
"valid_targets_min": 875
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.5632000000000001,
|
| 542 |
+
"grad_norm": 0.28723889701430083,
|
| 543 |
+
"learning_rate": 3.4847005892082266e-05,
|
| 544 |
+
"loss": 0.4086,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.10771428048610687,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 5469.1,
|
| 549 |
+
"valid_targets_min": 816
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.5952,
|
| 553 |
+
"grad_norm": 0.31601519650893833,
|
| 554 |
+
"learning_rate": 3.454521065542273e-05,
|
| 555 |
+
"loss": 0.4132,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.11569797992706299,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 4334.8,
|
| 560 |
+
"valid_targets_min": 688
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.6272,
|
| 564 |
+
"grad_norm": 0.35135975450130674,
|
| 565 |
+
"learning_rate": 3.423621541950597e-05,
|
| 566 |
+
"loss": 0.4265,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.1070767492055893,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 3814.1,
|
| 571 |
+
"valid_targets_min": 785
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.6592,
|
| 575 |
+
"grad_norm": 0.29723308168544715,
|
| 576 |
+
"learning_rate": 3.3920173139522664e-05,
|
| 577 |
+
"loss": 0.4188,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.10516910254955292,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 3706.8,
|
| 582 |
+
"valid_targets_min": 800
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.6912,
|
| 586 |
+
"grad_norm": 0.2768385567348046,
|
| 587 |
+
"learning_rate": 3.35972402590084e-05,
|
| 588 |
+
"loss": 0.4055,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.09208667278289795,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 3703.6,
|
| 593 |
+
"valid_targets_min": 550
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7231999999999998,
|
| 597 |
+
"grad_norm": 0.2951856918024355,
|
| 598 |
+
"learning_rate": 3.326757663240291e-05,
|
| 599 |
+
"loss": 0.4227,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.1036691889166832,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 3547.4,
|
| 604 |
+
"valid_targets_min": 713
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7551999999999999,
|
| 608 |
+
"grad_norm": 0.26249382923658837,
|
| 609 |
+
"learning_rate": 3.293134544592073e-05,
|
| 610 |
+
"loss": 0.4182,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.09299944341182709,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 4366.1,
|
| 615 |
+
"valid_targets_min": 794
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7872,
|
| 619 |
+
"grad_norm": 0.29210022832805477,
|
| 620 |
+
"learning_rate": 3.258871313677274e-05,
|
| 621 |
+
"loss": 0.3992,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.10157176852226257,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 3761.3,
|
| 626 |
+
"valid_targets_min": 696
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8192,
|
| 630 |
+
"grad_norm": 0.27904682672912556,
|
| 631 |
+
"learning_rate": 3.2239849310778316e-05,
|
| 632 |
+
"loss": 0.4214,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.09509153664112091,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 3494.4,
|
| 637 |
+
"valid_targets_min": 582
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.8512,
|
| 641 |
+
"grad_norm": 0.3252762284277571,
|
| 642 |
+
"learning_rate": 3.188492665840909e-05,
|
| 643 |
+
"loss": 0.4189,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.1019916832447052,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 3652.1,
|
| 648 |
+
"valid_targets_min": 851
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.8832,
|
| 652 |
+
"grad_norm": 0.2559216215832805,
|
| 653 |
+
"learning_rate": 3.1524120869305726e-05,
|
| 654 |
+
"loss": 0.4036,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.08774925768375397,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 3680.8,
|
| 659 |
+
"valid_targets_min": 785
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.9152,
|
| 663 |
+
"grad_norm": 0.28129794947346765,
|
| 664 |
+
"learning_rate": 3.11576105453101e-05,
|
| 665 |
+
"loss": 0.4012,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.10966770350933075,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 4628.2,
|
| 670 |
+
"valid_targets_min": 585
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.9472,
|
| 674 |
+
"grad_norm": 0.3261593902401628,
|
| 675 |
+
"learning_rate": 3.0785577112055916e-05,
|
| 676 |
+
"loss": 0.4036,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.10367356240749359,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 4310.1,
|
| 681 |
+
"valid_targets_min": 892
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9792,
|
| 685 |
+
"grad_norm": 0.2946580636329257,
|
| 686 |
+
"learning_rate": 3.040820472916153e-05,
|
| 687 |
+
"loss": 0.4056,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.10338163375854492,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 4015.2,
|
| 692 |
+
"valid_targets_min": 934
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.0064,
|
| 696 |
+
"grad_norm": 0.3039500754589427,
|
| 697 |
+
"learning_rate": 3.002568019906939e-05,
|
| 698 |
+
"loss": 0.4175,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.09754990041255951,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 3799.8,
|
| 703 |
+
"valid_targets_min": 754
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0384,
|
| 707 |
+
"grad_norm": 0.295008199316107,
|
| 708 |
+
"learning_rate": 2.963819287457733e-05,
|
| 709 |
+
"loss": 0.3919,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.09975887835025787,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 3814.1,
|
| 714 |
+
"valid_targets_min": 654
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0704,
|
| 718 |
+
"grad_norm": 0.28796119211965765,
|
| 719 |
+
"learning_rate": 2.924593456510733e-05,
|
| 720 |
+
"loss": 0.4054,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.09366752952337265,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 4051.3,
|
| 725 |
+
"valid_targets_min": 627
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.1024,
|
| 729 |
+
"grad_norm": 0.32513160009341713,
|
| 730 |
+
"learning_rate": 2.8849099441758306e-05,
|
| 731 |
+
"loss": 0.3913,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.10308220982551575,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 4733.2,
|
| 736 |
+
"valid_targets_min": 1213
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.1344,
|
| 740 |
+
"grad_norm": 0.24307889326670307,
|
| 741 |
+
"learning_rate": 2.844788394118979e-05,
|
| 742 |
+
"loss": 0.4093,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.08254360407590866,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 4415.9,
|
| 747 |
+
"valid_targets_min": 549
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.1664,
|
| 751 |
+
"grad_norm": 0.28397293492211034,
|
| 752 |
+
"learning_rate": 2.8042486668384164e-05,
|
| 753 |
+
"loss": 0.4046,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.11996494978666306,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 4177.9,
|
| 758 |
+
"valid_targets_min": 578
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.1984,
|
| 762 |
+
"grad_norm": 0.28612539051013547,
|
| 763 |
+
"learning_rate": 2.7633108298335582e-05,
|
| 764 |
+
"loss": 0.4116,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.1003495305776596,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 3604.4,
|
| 769 |
+
"valid_targets_min": 627
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.2304,
|
| 773 |
+
"grad_norm": 0.2901012203725845,
|
| 774 |
+
"learning_rate": 2.721995147671416e-05,
|
| 775 |
+
"loss": 0.4047,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.10313699394464493,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 4313.1,
|
| 780 |
+
"valid_targets_min": 830
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.2624,
|
| 784 |
+
"grad_norm": 0.339340824279522,
|
| 785 |
+
"learning_rate": 2.68032207195547e-05,
|
| 786 |
+
"loss": 0.4084,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.09717093408107758,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 2634.5,
|
| 791 |
+
"valid_targets_min": 622
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.2944,
|
| 795 |
+
"grad_norm": 0.3356431583635563,
|
| 796 |
+
"learning_rate": 2.6383122312019604e-05,
|
| 797 |
+
"loss": 0.3965,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.11940605938434601,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 4358.6,
|
| 802 |
+
"valid_targets_min": 653
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3264,
|
| 806 |
+
"grad_norm": 0.2729074469449499,
|
| 807 |
+
"learning_rate": 2.595986420628597e-05,
|
| 808 |
+
"loss": 0.3957,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.10392723232507706,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 4498.1,
|
| 813 |
+
"valid_targets_min": 1129
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3584,
|
| 817 |
+
"grad_norm": 0.2893023850029988,
|
| 818 |
+
"learning_rate": 2.5533655918607573e-05,
|
| 819 |
+
"loss": 0.3944,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.11088322103023529,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 4274.2,
|
| 824 |
+
"valid_targets_min": 593
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.3904,
|
| 828 |
+
"grad_norm": 0.25313367295678246,
|
| 829 |
+
"learning_rate": 2.510470842560259e-05,
|
| 830 |
+
"loss": 0.3795,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.09681592136621475,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 4609.8,
|
| 835 |
+
"valid_targets_min": 799
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4224,
|
| 839 |
+
"grad_norm": 0.2936301548074808,
|
| 840 |
+
"learning_rate": 2.467323405981841e-05,
|
| 841 |
+
"loss": 0.3992,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.10911795496940613,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 3933.9,
|
| 846 |
+
"valid_targets_min": 962
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.4544,
|
| 850 |
+
"grad_norm": 0.2731075417593816,
|
| 851 |
+
"learning_rate": 2.423944640462533e-05,
|
| 852 |
+
"loss": 0.4039,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.11098127067089081,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 4835.6,
|
| 857 |
+
"valid_targets_min": 587
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.4864,
|
| 861 |
+
"grad_norm": 0.291325200016876,
|
| 862 |
+
"learning_rate": 2.3803560188490968e-05,
|
| 863 |
+
"loss": 0.4048,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.09697733074426651,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 3457.2,
|
| 868 |
+
"valid_targets_min": 803
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.5183999999999997,
|
| 872 |
+
"grad_norm": 0.28348123924932705,
|
| 873 |
+
"learning_rate": 2.336579117868789e-05,
|
| 874 |
+
"loss": 0.3915,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.09868316352367401,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 4449.7,
|
| 879 |
+
"valid_targets_min": 571
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5504,
|
| 883 |
+
"grad_norm": 0.26609595339552466,
|
| 884 |
+
"learning_rate": 2.292635607448711e-05,
|
| 885 |
+
"loss": 0.3895,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.0807412639260292,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 3627.1,
|
| 890 |
+
"valid_targets_min": 661
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.5824,
|
| 894 |
+
"grad_norm": 0.33358859230174703,
|
| 895 |
+
"learning_rate": 2.248547239989008e-05,
|
| 896 |
+
"loss": 0.398,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.10890471935272217,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 3517.1,
|
| 901 |
+
"valid_targets_min": 585
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.6144,
|
| 905 |
+
"grad_norm": 0.3225642644827852,
|
| 906 |
+
"learning_rate": 2.204335839595255e-05,
|
| 907 |
+
"loss": 0.3952,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.10589894652366638,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 3461.2,
|
| 912 |
+
"valid_targets_min": 483
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6464,
|
| 916 |
+
"grad_norm": 0.28098105618711433,
|
| 917 |
+
"learning_rate": 2.1600232912753452e-05,
|
| 918 |
+
"loss": 0.3962,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.09055127203464508,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 3736.4,
|
| 923 |
+
"valid_targets_min": 919
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.6784,
|
| 927 |
+
"grad_norm": 0.28850076772839844,
|
| 928 |
+
"learning_rate": 2.1156315301062293e-05,
|
| 929 |
+
"loss": 0.3939,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.10580737888813019,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 3956.4,
|
| 934 |
+
"valid_targets_min": 840
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.7104,
|
| 938 |
+
"grad_norm": 0.2640927061351806,
|
| 939 |
+
"learning_rate": 2.0711825303758712e-05,
|
| 940 |
+
"loss": 0.4044,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.10874482244253159,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 4151.2,
|
| 945 |
+
"valid_targets_min": 734
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7424,
|
| 949 |
+
"grad_norm": 0.2870930699056917,
|
| 950 |
+
"learning_rate": 2.0266982947057962e-05,
|
| 951 |
+
"loss": 0.3872,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.08751390129327774,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 3402.9,
|
| 956 |
+
"valid_targets_min": 658
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.7744,
|
| 960 |
+
"grad_norm": 0.2676262206004521,
|
| 961 |
+
"learning_rate": 1.9822008431596083e-05,
|
| 962 |
+
"loss": 0.3992,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.10693050920963287,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 4426.9,
|
| 967 |
+
"valid_targets_min": 1543
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.8064,
|
| 971 |
+
"grad_norm": 0.2717783395966012,
|
| 972 |
+
"learning_rate": 1.937712202342881e-05,
|
| 973 |
+
"loss": 0.3901,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.10182879120111465,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 4275.3,
|
| 978 |
+
"valid_targets_min": 720
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8384,
|
| 982 |
+
"grad_norm": 0.28314115940486834,
|
| 983 |
+
"learning_rate": 1.8932543944998037e-05,
|
| 984 |
+
"loss": 0.4143,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.09646420925855637,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 4359.9,
|
| 989 |
+
"valid_targets_min": 585
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.8704,
|
| 993 |
+
"grad_norm": 0.27804663750003045,
|
| 994 |
+
"learning_rate": 1.8488494266119877e-05,
|
| 995 |
+
"loss": 0.3923,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.10060509294271469,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 3812.6,
|
| 1000 |
+
"valid_targets_min": 446
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.9024,
|
| 1004 |
+
"grad_norm": 0.28764089042619995,
|
| 1005 |
+
"learning_rate": 1.804519279504834e-05,
|
| 1006 |
+
"loss": 0.4051,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.10025319457054138,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 3917.0,
|
| 1011 |
+
"valid_targets_min": 558
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9344,
|
| 1015 |
+
"grad_norm": 0.2505864627102052,
|
| 1016 |
+
"learning_rate": 1.7602858969668365e-05,
|
| 1017 |
+
"loss": 0.3932,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.09689773619174957,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 4338.6,
|
| 1022 |
+
"valid_targets_min": 1069
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.9664,
|
| 1026 |
+
"grad_norm": 0.23037767347131669,
|
| 1027 |
+
"learning_rate": 1.716171174887231e-05,
|
| 1028 |
+
"loss": 0.3841,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.09673911333084106,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 4582.5,
|
| 1033 |
+
"valid_targets_min": 810
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.9984,
|
| 1037 |
+
"grad_norm": 0.27508918041338687,
|
| 1038 |
+
"learning_rate": 1.6721969504173484e-05,
|
| 1039 |
+
"loss": 0.3979,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.09660154581069946,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 3506.7,
|
| 1044 |
+
"valid_targets_min": 618
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0256,
|
| 1048 |
+
"grad_norm": 0.278832581896705,
|
| 1049 |
+
"learning_rate": 1.628384991161041e-05,
|
| 1050 |
+
"loss": 0.3801,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.09370477497577667,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 3882.2,
|
| 1055 |
+
"valid_targets_min": 999
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.0576,
|
| 1059 |
+
"grad_norm": 0.27988835500157977,
|
| 1060 |
+
"learning_rate": 1.5847569843995452e-05,
|
| 1061 |
+
"loss": 0.3895,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.1039055809378624,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 3686.8,
|
| 1066 |
+
"valid_targets_min": 767
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.0896,
|
| 1070 |
+
"grad_norm": 0.27300541572988496,
|
| 1071 |
+
"learning_rate": 1.5413345263560922e-05,
|
| 1072 |
+
"loss": 0.392,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.11811760067939758,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 5150.7,
|
| 1077 |
+
"valid_targets_min": 1112
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.1216,
|
| 1081 |
+
"grad_norm": 0.2898076648663478,
|
| 1082 |
+
"learning_rate": 1.4981391115056032e-05,
|
| 1083 |
+
"loss": 0.3881,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.11502294987440109,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 4739.1,
|
| 1088 |
+
"valid_targets_min": 1357
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1536,
|
| 1092 |
+
"grad_norm": 0.2820293732275994,
|
| 1093 |
+
"learning_rate": 1.455192121934748e-05,
|
| 1094 |
+
"loss": 0.3796,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.09026939421892166,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 3438.8,
|
| 1099 |
+
"valid_targets_min": 613
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.1856,
|
| 1103 |
+
"grad_norm": 0.2563623556654072,
|
| 1104 |
+
"learning_rate": 1.4125148167576303e-05,
|
| 1105 |
+
"loss": 0.3891,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.07436763495206833,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 3468.8,
|
| 1110 |
+
"valid_targets_min": 626
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.2176,
|
| 1114 |
+
"grad_norm": 0.2639018773325639,
|
| 1115 |
+
"learning_rate": 1.3701283215923563e-05,
|
| 1116 |
+
"loss": 0.3872,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.0847480371594429,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 3150.1,
|
| 1121 |
+
"valid_targets_min": 720
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2496,
|
| 1125 |
+
"grad_norm": 0.28414234275420414,
|
| 1126 |
+
"learning_rate": 1.328053618103677e-05,
|
| 1127 |
+
"loss": 0.3972,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.11753346771001816,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 4629.7,
|
| 1132 |
+
"valid_targets_min": 1059
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.2816,
|
| 1136 |
+
"grad_norm": 0.24480188835818936,
|
| 1137 |
+
"learning_rate": 1.2863115336168916e-05,
|
| 1138 |
+
"loss": 0.3837,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.09535688906908035,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 4956.4,
|
| 1143 |
+
"valid_targets_min": 724
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3136,
|
| 1147 |
+
"grad_norm": 0.27121801267241674,
|
| 1148 |
+
"learning_rate": 1.2449227308081509e-05,
|
| 1149 |
+
"loss": 0.3931,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.10601663589477539,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 4022.4,
|
| 1154 |
+
"valid_targets_min": 1571
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.3456,
|
| 1158 |
+
"grad_norm": 0.2728948280882927,
|
| 1159 |
+
"learning_rate": 1.2039076974762587e-05,
|
| 1160 |
+
"loss": 0.3947,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.10746031999588013,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 3813.9,
|
| 1165 |
+
"valid_targets_min": 887
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.3776,
|
| 1169 |
+
"grad_norm": 0.28856355078117213,
|
| 1170 |
+
"learning_rate": 1.163286736401044e-05,
|
| 1171 |
+
"loss": 0.3869,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.11251718550920486,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 3836.1,
|
| 1176 |
+
"valid_targets_min": 1244
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4096,
|
| 1180 |
+
"grad_norm": 0.2660317815137939,
|
| 1181 |
+
"learning_rate": 1.123079955293322e-05,
|
| 1182 |
+
"loss": 0.384,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.0875939428806305,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 3313.4,
|
| 1187 |
+
"valid_targets_min": 923
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.4416,
|
| 1191 |
+
"grad_norm": 0.3266588391926982,
|
| 1192 |
+
"learning_rate": 1.0833072568414037e-05,
|
| 1193 |
+
"loss": 0.3959,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.1070742979645729,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 4143.3,
|
| 1198 |
+
"valid_targets_min": 1343
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.4736000000000002,
|
| 1202 |
+
"grad_norm": 0.25227560229994916,
|
| 1203 |
+
"learning_rate": 1.0439883288591057e-05,
|
| 1204 |
+
"loss": 0.3735,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.09336927533149719,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 4148.4,
|
| 1209 |
+
"valid_targets_min": 627
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.5056000000000003,
|
| 1213 |
+
"grad_norm": 0.29732898476146347,
|
| 1214 |
+
"learning_rate": 1.0051426345401202e-05,
|
| 1215 |
+
"loss": 0.3873,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.09397479891777039,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 3452.2,
|
| 1220 |
+
"valid_targets_min": 635
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5376,
|
| 1224 |
+
"grad_norm": 0.28103656307884795,
|
| 1225 |
+
"learning_rate": 9.667894028235704e-06,
|
| 1226 |
+
"loss": 0.3916,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.0950830802321434,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 3525.2,
|
| 1231 |
+
"valid_targets_min": 874
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.5696,
|
| 1235 |
+
"grad_norm": 0.27515789289270587,
|
| 1236 |
+
"learning_rate": 9.289476188755315e-06,
|
| 1237 |
+
"loss": 0.3873,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.08731421828269958,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 3457.8,
|
| 1242 |
+
"valid_targets_min": 816
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6016,
|
| 1246 |
+
"grad_norm": 0.26595925071142734,
|
| 1247 |
+
"learning_rate": 8.916360146912122e-06,
|
| 1248 |
+
"loss": 0.3903,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.0943126454949379,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 3755.8,
|
| 1253 |
+
"valid_targets_min": 916
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.6336,
|
| 1257 |
+
"grad_norm": 0.26908934868333345,
|
| 1258 |
+
"learning_rate": 8.548730598224646e-06,
|
| 1259 |
+
"loss": 0.3881,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.11741252988576889,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 4463.9,
|
| 1264 |
+
"valid_targets_min": 1490
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.6656,
|
| 1268 |
+
"grad_norm": 0.26163071888829387,
|
| 1269 |
+
"learning_rate": 8.186769522352053e-06,
|
| 1270 |
+
"loss": 0.3804,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.07929997891187668,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 2767.7,
|
| 1275 |
+
"valid_targets_min": 668
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.6976,
|
| 1279 |
+
"grad_norm": 0.2762268133622116,
|
| 1280 |
+
"learning_rate": 7.830656093012714e-06,
|
| 1281 |
+
"loss": 0.3876,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.08485132455825806,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 3037.7,
|
| 1286 |
+
"valid_targets_min": 726
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.7296,
|
| 1290 |
+
"grad_norm": 0.2674780145990845,
|
| 1291 |
+
"learning_rate": 7.480566589291696e-06,
|
| 1292 |
+
"loss": 0.3872,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.09651274234056473,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 4118.8,
|
| 1297 |
+
"valid_targets_min": 1165
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.7616,
|
| 1301 |
+
"grad_norm": 0.24858944575790964,
|
| 1302 |
+
"learning_rate": 7.1366743083812285e-06,
|
| 1303 |
+
"loss": 0.3843,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.09582122415304184,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 4209.6,
|
| 1308 |
+
"valid_targets_min": 790
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.7936,
|
| 1312 |
+
"grad_norm": 0.26046610758966066,
|
| 1313 |
+
"learning_rate": 6.799149479797101e-06,
|
| 1314 |
+
"loss": 0.3849,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.08760522305965424,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 3850.9,
|
| 1319 |
+
"valid_targets_min": 844
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.8256,
|
| 1323 |
+
"grad_norm": 0.2517690965363384,
|
| 1324 |
+
"learning_rate": 6.4681591811137e-06,
|
| 1325 |
+
"loss": 0.3862,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.10714523494243622,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 5318.4,
|
| 1330 |
+
"valid_targets_min": 1305
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.8576,
|
| 1334 |
+
"grad_norm": 0.27076985537219617,
|
| 1335 |
+
"learning_rate": 6.143867255259197e-06,
|
| 1336 |
+
"loss": 0.3801,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.09001049399375916,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 3394.0,
|
| 1341 |
+
"valid_targets_min": 991
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.8895999999999997,
|
| 1345 |
+
"grad_norm": 0.2825297290428291,
|
| 1346 |
+
"learning_rate": 5.8264342294119504e-06,
|
| 1347 |
+
"loss": 0.398,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.09313380718231201,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 3341.5,
|
| 1352 |
+
"valid_targets_min": 788
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.9215999999999998,
|
| 1356 |
+
"grad_norm": 0.26545079772678515,
|
| 1357 |
+
"learning_rate": 5.516017235538258e-06,
|
| 1358 |
+
"loss": 0.3819,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.08457700908184052,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 3106.5,
|
| 1363 |
+
"valid_targets_min": 660
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.9536,
|
| 1367 |
+
"grad_norm": 0.2567709880978487,
|
| 1368 |
+
"learning_rate": 5.212769932610695e-06,
|
| 1369 |
+
"loss": 0.3847,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.08885794878005981,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 4086.2,
|
| 1374 |
+
"valid_targets_min": 743
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.9856,
|
| 1378 |
+
"grad_norm": 0.24894909802272616,
|
| 1379 |
+
"learning_rate": 4.916842430545681e-06,
|
| 1380 |
+
"loss": 0.3895,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.0957787036895752,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 4743.1,
|
| 1385 |
+
"valid_targets_min": 934
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.0128,
|
| 1389 |
+
"grad_norm": 0.2730521094355628,
|
| 1390 |
+
"learning_rate": 4.628381215897837e-06,
|
| 1391 |
+
"loss": 0.3959,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.09598704427480698,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 3783.9,
|
| 1396 |
+
"valid_targets_min": 1002
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.0448,
|
| 1400 |
+
"grad_norm": 0.2517076111149134,
|
| 1401 |
+
"learning_rate": 4.347529079347914e-06,
|
| 1402 |
+
"loss": 0.375,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.10451523959636688,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 4389.2,
|
| 1407 |
+
"valid_targets_min": 764
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.0768,
|
| 1411 |
+
"grad_norm": 0.26123311146967465,
|
| 1412 |
+
"learning_rate": 4.074425045020247e-06,
|
| 1413 |
+
"loss": 0.3842,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.085345059633255,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 3496.9,
|
| 1418 |
+
"valid_targets_min": 1141
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.1088,
|
| 1422 |
+
"grad_norm": 0.28008604571769974,
|
| 1423 |
+
"learning_rate": 3.8092043016646487e-06,
|
| 1424 |
+
"loss": 0.3849,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.10877778381109238,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 3883.4,
|
| 1429 |
+
"valid_targets_min": 836
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.1408,
|
| 1433 |
+
"grad_norm": 0.2810668056658052,
|
| 1434 |
+
"learning_rate": 3.551998135736867e-06,
|
| 1435 |
+
"loss": 0.3942,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.103969506919384,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 4210.5,
|
| 1440 |
+
"valid_targets_min": 1079
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.1728,
|
| 1444 |
+
"grad_norm": 0.2679325409425069,
|
| 1445 |
+
"learning_rate": 3.3029338664107267e-06,
|
| 1446 |
+
"loss": 0.3697,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.10508221387863159,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 3703.6,
|
| 1451 |
+
"valid_targets_min": 1064
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.2048,
|
| 1455 |
+
"grad_norm": 0.2677720958501848,
|
| 1456 |
+
"learning_rate": 3.0621347825540625e-06,
|
| 1457 |
+
"loss": 0.3743,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.09485597908496857,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 3589.9,
|
| 1462 |
+
"valid_targets_min": 752
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.2368,
|
| 1466 |
+
"grad_norm": 0.25540762686442,
|
| 1467 |
+
"learning_rate": 2.8297200816997183e-06,
|
| 1468 |
+
"loss": 0.3828,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.08198321610689163,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 3685.8,
|
| 1473 |
+
"valid_targets_min": 587
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.2688,
|
| 1477 |
+
"grad_norm": 0.2538363359478502,
|
| 1478 |
+
"learning_rate": 2.605804811041803e-06,
|
| 1479 |
+
"loss": 0.3829,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.10491339862346649,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 4984.3,
|
| 1484 |
+
"valid_targets_min": 698
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.3008,
|
| 1488 |
+
"grad_norm": 0.24929527911049837,
|
| 1489 |
+
"learning_rate": 2.390499810486351e-06,
|
| 1490 |
+
"loss": 0.3805,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.08019746094942093,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 3777.9,
|
| 1495 |
+
"valid_targets_min": 680
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.3328,
|
| 1499 |
+
"grad_norm": 0.2688823154765233,
|
| 1500 |
+
"learning_rate": 2.183911657784685e-06,
|
| 1501 |
+
"loss": 0.3892,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.11593001335859299,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 4966.8,
|
| 1506 |
+
"valid_targets_min": 1682
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.3648,
|
| 1510 |
+
"grad_norm": 0.2681893697414637,
|
| 1511 |
+
"learning_rate": 1.986142615776532e-06,
|
| 1512 |
+
"loss": 0.3824,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.09049779176712036,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 3786.8,
|
| 1517 |
+
"valid_targets_min": 553
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.3968,
|
| 1521 |
+
"grad_norm": 0.25803539534975506,
|
| 1522 |
+
"learning_rate": 1.7972905817690644e-06,
|
| 1523 |
+
"loss": 0.3822,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.09479928016662598,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 4407.2,
|
| 1528 |
+
"valid_targets_min": 701
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.4288,
|
| 1532 |
+
"grad_norm": 0.2721781700256369,
|
| 1533 |
+
"learning_rate": 1.617449039076955e-06,
|
| 1534 |
+
"loss": 0.3817,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.09600777924060822,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 4190.6,
|
| 1539 |
+
"valid_targets_min": 707
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.4608,
|
| 1543 |
+
"grad_norm": 0.27203666956952044,
|
| 1544 |
+
"learning_rate": 1.4467070107473413e-06,
|
| 1545 |
+
"loss": 0.3936,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.1137179434299469,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 4398.9,
|
| 1550 |
+
"valid_targets_min": 2071
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.4928,
|
| 1554 |
+
"grad_norm": 0.28700859136150275,
|
| 1555 |
+
"learning_rate": 1.2851490154926816e-06,
|
| 1556 |
+
"loss": 0.3863,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.0846642404794693,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 3038.7,
|
| 1561 |
+
"valid_targets_min": 724
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.5248,
|
| 1565 |
+
"grad_norm": 0.2529565401900442,
|
| 1566 |
+
"learning_rate": 1.1328550258533211e-06,
|
| 1567 |
+
"loss": 0.377,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.10368967056274414,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 4356.4,
|
| 1572 |
+
"valid_targets_min": 904
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.5568,
|
| 1576 |
+
"grad_norm": 0.2500059167539887,
|
| 1577 |
+
"learning_rate": 9.899004286103953e-07,
|
| 1578 |
+
"loss": 0.3834,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.08508186787366867,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 3736.4,
|
| 1583 |
+
"valid_targets_min": 788
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.5888,
|
| 1587 |
+
"grad_norm": 0.27361014801466704,
|
| 1588 |
+
"learning_rate": 8.5635598746876e-07,
|
| 1589 |
+
"loss": 0.3909,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.09339834749698639,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 3070.9,
|
| 1594 |
+
"valid_targets_min": 523
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.6208,
|
| 1598 |
+
"grad_norm": 0.25039458600858977,
|
| 1599 |
+
"learning_rate": 7.32287808028389e-07,
|
| 1600 |
+
"loss": 0.3735,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.07535479962825775,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 3382.6,
|
| 1605 |
+
"valid_targets_min": 794
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.6528,
|
| 1609 |
+
"grad_norm": 0.2846141316902581,
|
| 1610 |
+
"learning_rate": 6.177573050615327e-07,
|
| 1611 |
+
"loss": 0.3891,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.08932908624410629,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 3472.3,
|
| 1616 |
+
"valid_targets_min": 806
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.6848,
|
| 1620 |
+
"grad_norm": 0.25547731029216575,
|
| 1621 |
+
"learning_rate": 5.128211721119213e-07,
|
| 1622 |
+
"loss": 0.3894,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.09996099770069122,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 4633.1,
|
| 1627 |
+
"valid_targets_min": 1030
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.7168,
|
| 1631 |
+
"grad_norm": 0.25171677147508753,
|
| 1632 |
+
"learning_rate": 4.175313534309755e-07,
|
| 1633 |
+
"loss": 0.3803,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.08795123547315598,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 3785.0,
|
| 1638 |
+
"valid_targets_min": 545
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.7488,
|
| 1642 |
+
"grad_norm": 0.24110208408904205,
|
| 1643 |
+
"learning_rate": 3.319350182649861e-07,
|
| 1644 |
+
"loss": 0.39,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.10249342024326324,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 4599.1,
|
| 1649 |
+
"valid_targets_min": 1465
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.7808,
|
| 1653 |
+
"grad_norm": 0.272490256449447,
|
| 1654 |
+
"learning_rate": 2.560745375059392e-07,
|
| 1655 |
+
"loss": 0.3892,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.09520266950130463,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 3378.0,
|
| 1660 |
+
"valid_targets_min": 551
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.8128,
|
| 1664 |
+
"grad_norm": 0.27541100522754647,
|
| 1665 |
+
"learning_rate": 1.8998746271758016e-07,
|
| 1666 |
+
"loss": 0.389,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.09170030802488327,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 4346.1,
|
| 1671 |
+
"valid_targets_min": 1001
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.8448,
|
| 1675 |
+
"grad_norm": 0.24879985587571116,
|
| 1676 |
+
"learning_rate": 1.337065075470778e-07,
|
| 1677 |
+
"loss": 0.3775,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.08432689309120178,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 3557.9,
|
| 1682 |
+
"valid_targets_min": 481
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.8768,
|
| 1686 |
+
"grad_norm": 0.25501966207119003,
|
| 1687 |
+
"learning_rate": 8.725953153150279e-08,
|
| 1688 |
+
"loss": 0.3674,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.09955944120883942,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 4503.8,
|
| 1693 |
+
"valid_targets_min": 1642
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.9088,
|
| 1697 |
+
"grad_norm": 0.24823600991634784,
|
| 1698 |
+
"learning_rate": 5.066952630711886e-08,
|
| 1699 |
+
"loss": 0.3811,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.08799436688423157,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 4153.3,
|
| 1704 |
+
"valid_targets_min": 703
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.9408,
|
| 1708 |
+
"grad_norm": 0.2925007672229399,
|
| 1709 |
+
"learning_rate": 2.3954604228342283e-08,
|
| 1710 |
+
"loss": 0.3864,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.10177664458751678,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 4148.4,
|
| 1715 |
+
"valid_targets_min": 805
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.9728,
|
| 1719 |
+
"grad_norm": 0.27129018782269604,
|
| 1720 |
+
"learning_rate": 7.12798940197601e-09,
|
| 1721 |
+
"loss": 0.3899,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.10262542217969894,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 4038.1,
|
| 1726 |
+
"valid_targets_min": 1461
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"grad_norm": 0.5130672799557714,
|
| 1731 |
+
"learning_rate": 1.9801114115480802e-10,
|
| 1732 |
+
"loss": 0.3782,
|
| 1733 |
+
"loss_nan_ranks": 0,
|
| 1734 |
+
"loss_rank_avg": 0.40260475873947144,
|
| 1735 |
+
"step": 785,
|
| 1736 |
+
"valid_targets_mean": 3870.8,
|
| 1737 |
+
"valid_targets_min": 1612
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 5.0,
|
| 1741 |
+
"loss_nan_ranks": 0,
|
| 1742 |
+
"loss_rank_avg": 0.40260475873947144,
|
| 1743 |
+
"step": 785,
|
| 1744 |
+
"total_flos": 1.3461651925678162e+18,
|
| 1745 |
+
"train_loss": 0.41955339954157544,
|
| 1746 |
+
"train_runtime": 17506.3111,
|
| 1747 |
+
"train_samples_per_second": 2.856,
|
| 1748 |
+
"train_steps_per_second": 0.045,
|
| 1749 |
+
"valid_targets_mean": 3870.8,
|
| 1750 |
+
"valid_targets_min": 1612
|
| 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": 1.3461651925678162e+18,
|
| 1771 |
+
"train_batch_size": 1,
|
| 1772 |
+
"trial_name": null,
|
| 1773 |
+
"trial_params": null
|
| 1774 |
+
}
|
training_loss.png
ADDED
|