End of training
Browse files- README.md +2 -1
- all_results.json +16 -0
- train_results.json +16 -0
- trainer_state.json +1763 -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: codeforces-sandboxes-traces-terminus-2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# codeforces-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: codeforces-sandboxes-traces-terminus-2
|
|
|
|
| 16 |
|
| 17 |
# codeforces-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/codeforces-sandboxes-traces-terminus-2 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.606449803467651,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 189.7226555180974,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.11478559672832489,
|
| 7 |
+
"mfu_percent": 0.3255441557220955,
|
| 8 |
+
"mfu_percent_theoretical": 13.407961520713599,
|
| 9 |
+
"total_flos": 1.0263973305862062e+18,
|
| 10 |
+
"train_loss": 0.4329962073228298,
|
| 11 |
+
"train_runtime": 13926.0897,
|
| 12 |
+
"train_samples_per_second": 3.575,
|
| 13 |
+
"train_steps_per_second": 0.056,
|
| 14 |
+
"valid_targets_mean": 3291.8,
|
| 15 |
+
"valid_targets_min": 500
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.606449803467651,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 189.7226555180974,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.11478559672832489,
|
| 7 |
+
"mfu_percent": 0.3255441557220955,
|
| 8 |
+
"mfu_percent_theoretical": 13.407961520713599,
|
| 9 |
+
"total_flos": 1.0263973305862062e+18,
|
| 10 |
+
"train_loss": 0.4329962073228298,
|
| 11 |
+
"train_runtime": 13926.0897,
|
| 12 |
+
"train_samples_per_second": 3.575,
|
| 13 |
+
"train_steps_per_second": 0.056,
|
| 14 |
+
"valid_targets_mean": 3291.8,
|
| 15 |
+
"valid_targets_min": 500
|
| 16 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1763 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 780,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03210272873194221,
|
| 14 |
+
"grad_norm": 7.443076512121371,
|
| 15 |
+
"learning_rate": 2.0512820512820513e-06,
|
| 16 |
+
"loss": 0.8032,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.21804624795913696,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 4336.2,
|
| 21 |
+
"valid_targets_min": 516
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.06420545746388442,
|
| 25 |
+
"grad_norm": 5.001718673683824,
|
| 26 |
+
"learning_rate": 4.615384615384616e-06,
|
| 27 |
+
"loss": 0.7807,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.21758630871772766,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 3736.1,
|
| 32 |
+
"valid_targets_min": 1113
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.09630818619582665,
|
| 36 |
+
"grad_norm": 2.9021650531373258,
|
| 37 |
+
"learning_rate": 7.17948717948718e-06,
|
| 38 |
+
"loss": 0.7119,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.18503277003765106,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 3926.9,
|
| 43 |
+
"valid_targets_min": 837
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.12841091492776885,
|
| 47 |
+
"grad_norm": 1.0697186254007953,
|
| 48 |
+
"learning_rate": 9.743589743589744e-06,
|
| 49 |
+
"loss": 0.6496,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.15346118807792664,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 3225.6,
|
| 54 |
+
"valid_targets_min": 613
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.16051364365971107,
|
| 58 |
+
"grad_norm": 0.9133287345600813,
|
| 59 |
+
"learning_rate": 1.230769230769231e-05,
|
| 60 |
+
"loss": 0.6397,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.16350162029266357,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 3446.9,
|
| 65 |
+
"valid_targets_min": 1106
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1926163723916533,
|
| 69 |
+
"grad_norm": 0.5525491725186413,
|
| 70 |
+
"learning_rate": 1.4871794871794874e-05,
|
| 71 |
+
"loss": 0.6154,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.16575123369693756,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 4915.8,
|
| 76 |
+
"valid_targets_min": 1045
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.2247191011235955,
|
| 80 |
+
"grad_norm": 0.5368849244599112,
|
| 81 |
+
"learning_rate": 1.7435897435897438e-05,
|
| 82 |
+
"loss": 0.5911,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.14013072848320007,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 3104.2,
|
| 87 |
+
"valid_targets_min": 859
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2568218298555377,
|
| 91 |
+
"grad_norm": 0.4615281982796634,
|
| 92 |
+
"learning_rate": 2e-05,
|
| 93 |
+
"loss": 0.5892,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.161387100815773,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 3632.9,
|
| 98 |
+
"valid_targets_min": 659
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.28892455858747995,
|
| 102 |
+
"grad_norm": 0.4061270737531561,
|
| 103 |
+
"learning_rate": 2.2564102564102566e-05,
|
| 104 |
+
"loss": 0.5387,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.12181757390499115,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 3131.7,
|
| 109 |
+
"valid_targets_min": 505
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32102728731942215,
|
| 113 |
+
"grad_norm": 0.33878739727162444,
|
| 114 |
+
"learning_rate": 2.512820512820513e-05,
|
| 115 |
+
"loss": 0.5399,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.12523899972438812,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 3602.1,
|
| 120 |
+
"valid_targets_min": 609
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.35313001605136435,
|
| 124 |
+
"grad_norm": 0.35224338467065713,
|
| 125 |
+
"learning_rate": 2.7692307692307694e-05,
|
| 126 |
+
"loss": 0.5154,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.15735869109630585,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 4276.2,
|
| 131 |
+
"valid_targets_min": 1111
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.3852327447833066,
|
| 135 |
+
"grad_norm": 0.29204719939595825,
|
| 136 |
+
"learning_rate": 3.0256410256410257e-05,
|
| 137 |
+
"loss": 0.505,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.142252117395401,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 4205.7,
|
| 142 |
+
"valid_targets_min": 1221
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4173354735152488,
|
| 146 |
+
"grad_norm": 0.3108436023652963,
|
| 147 |
+
"learning_rate": 3.282051282051282e-05,
|
| 148 |
+
"loss": 0.5088,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.10603222995996475,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 3141.0,
|
| 153 |
+
"valid_targets_min": 1008
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.449438202247191,
|
| 157 |
+
"grad_norm": 0.2986968060504965,
|
| 158 |
+
"learning_rate": 3.538461538461539e-05,
|
| 159 |
+
"loss": 0.495,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.1236981451511383,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 3417.8,
|
| 164 |
+
"valid_targets_min": 816
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48154093097913325,
|
| 168 |
+
"grad_norm": 0.6323281716016163,
|
| 169 |
+
"learning_rate": 3.794871794871795e-05,
|
| 170 |
+
"loss": 0.5013,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.11899252235889435,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 3896.2,
|
| 175 |
+
"valid_targets_min": 749
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.5136436597110754,
|
| 179 |
+
"grad_norm": 0.3574090578830505,
|
| 180 |
+
"learning_rate": 3.999979972589652e-05,
|
| 181 |
+
"loss": 0.4691,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.12434444576501846,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 3266.9,
|
| 186 |
+
"valid_targets_min": 726
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.5457463884430177,
|
| 190 |
+
"grad_norm": 0.3233297771547147,
|
| 191 |
+
"learning_rate": 3.999279055341771e-05,
|
| 192 |
+
"loss": 0.4792,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.13975250720977783,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 4060.8,
|
| 197 |
+
"valid_targets_min": 1022
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.5778491171749599,
|
| 201 |
+
"grad_norm": 0.3481282366313697,
|
| 202 |
+
"learning_rate": 3.9975771686375434e-05,
|
| 203 |
+
"loss": 0.4805,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.07897774875164032,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 2103.4,
|
| 208 |
+
"valid_targets_min": 631
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.6099518459069021,
|
| 212 |
+
"grad_norm": 0.342138756435365,
|
| 213 |
+
"learning_rate": 3.9948751645524235e-05,
|
| 214 |
+
"loss": 0.4749,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.11953769624233246,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 3517.6,
|
| 219 |
+
"valid_targets_min": 1496
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6420545746388443,
|
| 223 |
+
"grad_norm": 0.33501528178098466,
|
| 224 |
+
"learning_rate": 3.9911743958858376e-05,
|
| 225 |
+
"loss": 0.4818,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.11692598462104797,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 3028.9,
|
| 230 |
+
"valid_targets_min": 1292
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.6741573033707865,
|
| 234 |
+
"grad_norm": 0.3030959299753308,
|
| 235 |
+
"learning_rate": 3.9864767154838864e-05,
|
| 236 |
+
"loss": 0.4759,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.08811567723751068,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 2688.0,
|
| 241 |
+
"valid_targets_min": 720
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.7062600321027287,
|
| 245 |
+
"grad_norm": 0.2891268278078363,
|
| 246 |
+
"learning_rate": 3.9807844753116864e-05,
|
| 247 |
+
"loss": 0.4494,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.10534675419330597,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 3910.9,
|
| 252 |
+
"valid_targets_min": 1069
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.7383627608346709,
|
| 256 |
+
"grad_norm": 0.27088323808017967,
|
| 257 |
+
"learning_rate": 3.9741005252758255e-05,
|
| 258 |
+
"loss": 0.4684,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.09811793267726898,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 3091.4,
|
| 263 |
+
"valid_targets_min": 1057
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.7704654895666132,
|
| 267 |
+
"grad_norm": 0.3246822966924674,
|
| 268 |
+
"learning_rate": 3.966428211797514e-05,
|
| 269 |
+
"loss": 0.4527,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.11549907922744751,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 3803.6,
|
| 274 |
+
"valid_targets_min": 992
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8025682182985554,
|
| 278 |
+
"grad_norm": 0.33072065469680983,
|
| 279 |
+
"learning_rate": 3.957771376137144e-05,
|
| 280 |
+
"loss": 0.4683,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.1350668966770172,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 3599.0,
|
| 285 |
+
"valid_targets_min": 579
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.8346709470304976,
|
| 289 |
+
"grad_norm": 0.35035972572140217,
|
| 290 |
+
"learning_rate": 3.9481343524711095e-05,
|
| 291 |
+
"loss": 0.4594,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.12366610765457153,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 3637.1,
|
| 296 |
+
"valid_targets_min": 780
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8667736757624398,
|
| 300 |
+
"grad_norm": 0.33737136455082983,
|
| 301 |
+
"learning_rate": 3.937521965721831e-05,
|
| 302 |
+
"loss": 0.4599,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.11112713068723679,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 4047.2,
|
| 307 |
+
"valid_targets_min": 1496
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.898876404494382,
|
| 311 |
+
"grad_norm": 0.31265257576082217,
|
| 312 |
+
"learning_rate": 3.9259395291420864e-05,
|
| 313 |
+
"loss": 0.4346,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.1048852801322937,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 3647.4,
|
| 318 |
+
"valid_targets_min": 1385
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.9309791332263242,
|
| 322 |
+
"grad_norm": 0.3107204486324316,
|
| 323 |
+
"learning_rate": 3.913392841654851e-05,
|
| 324 |
+
"loss": 0.452,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.13349255919456482,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 4499.4,
|
| 329 |
+
"valid_targets_min": 791
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.9630818619582665,
|
| 333 |
+
"grad_norm": 0.31387648518622074,
|
| 334 |
+
"learning_rate": 3.8998881849499756e-05,
|
| 335 |
+
"loss": 0.4438,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.09112104773521423,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 3127.7,
|
| 340 |
+
"valid_targets_min": 827
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.9951845906902087,
|
| 344 |
+
"grad_norm": 0.31502159646866623,
|
| 345 |
+
"learning_rate": 3.885432320339167e-05,
|
| 346 |
+
"loss": 0.4491,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.10973098129034042,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 3247.8,
|
| 351 |
+
"valid_targets_min": 860
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0256821829855538,
|
| 355 |
+
"grad_norm": 0.2995227413933546,
|
| 356 |
+
"learning_rate": 3.8700324853708304e-05,
|
| 357 |
+
"loss": 0.4309,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.12735167145729065,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 4369.8,
|
| 362 |
+
"valid_targets_min": 1151
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.057784911717496,
|
| 366 |
+
"grad_norm": 0.3265629724457608,
|
| 367 |
+
"learning_rate": 3.853696390206484e-05,
|
| 368 |
+
"loss": 0.4441,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.10513167083263397,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 3461.3,
|
| 373 |
+
"valid_targets_min": 1352
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0898876404494382,
|
| 377 |
+
"grad_norm": 0.32739344706824414,
|
| 378 |
+
"learning_rate": 3.8364322137605484e-05,
|
| 379 |
+
"loss": 0.4472,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.09724493324756622,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 3163.2,
|
| 384 |
+
"valid_targets_min": 806
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1219903691813804,
|
| 388 |
+
"grad_norm": 0.3171489222363104,
|
| 389 |
+
"learning_rate": 3.818248599605448e-05,
|
| 390 |
+
"loss": 0.4197,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.1121516227722168,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 4971.7,
|
| 395 |
+
"valid_targets_min": 925
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1540930979133226,
|
| 399 |
+
"grad_norm": 0.3273324285042093,
|
| 400 |
+
"learning_rate": 3.799154651644075e-05,
|
| 401 |
+
"loss": 0.4474,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.11072337627410889,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 3005.9,
|
| 406 |
+
"valid_targets_min": 1236
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1861958266452648,
|
| 410 |
+
"grad_norm": 0.28545216130602724,
|
| 411 |
+
"learning_rate": 3.7791599295517825e-05,
|
| 412 |
+
"loss": 0.4382,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.09122270345687866,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 4056.4,
|
| 417 |
+
"valid_targets_min": 941
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.218298555377207,
|
| 421 |
+
"grad_norm": 0.35115623987576444,
|
| 422 |
+
"learning_rate": 3.75827444399019e-05,
|
| 423 |
+
"loss": 0.4368,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.12966161966323853,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 4549.7,
|
| 428 |
+
"valid_targets_min": 1229
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.2504012841091492,
|
| 432 |
+
"grad_norm": 0.3024907929000345,
|
| 433 |
+
"learning_rate": 3.736508651595188e-05,
|
| 434 |
+
"loss": 0.4281,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.09971387684345245,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 3831.5,
|
| 439 |
+
"valid_targets_min": 1330
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2825040128410916,
|
| 443 |
+
"grad_norm": 0.3649481469747776,
|
| 444 |
+
"learning_rate": 3.713873449741669e-05,
|
| 445 |
+
"loss": 0.4343,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.12230262160301208,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 3829.6,
|
| 450 |
+
"valid_targets_min": 517
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3146067415730336,
|
| 454 |
+
"grad_norm": 0.3081241454749465,
|
| 455 |
+
"learning_rate": 3.69038017108759e-05,
|
| 456 |
+
"loss": 0.4231,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.09591006487607956,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 3110.9,
|
| 461 |
+
"valid_targets_min": 1204
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.346709470304976,
|
| 465 |
+
"grad_norm": 0.29545465896778206,
|
| 466 |
+
"learning_rate": 3.6660405779001036e-05,
|
| 467 |
+
"loss": 0.4419,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.08623912930488586,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 3680.1,
|
| 472 |
+
"valid_targets_min": 796
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3788121990369182,
|
| 476 |
+
"grad_norm": 0.37785862300967726,
|
| 477 |
+
"learning_rate": 3.640866856166601e-05,
|
| 478 |
+
"loss": 0.4384,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.10427422821521759,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 3281.7,
|
| 483 |
+
"valid_targets_min": 837
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4109149277688604,
|
| 487 |
+
"grad_norm": 0.30674701618911177,
|
| 488 |
+
"learning_rate": 3.614871609493614e-05,
|
| 489 |
+
"loss": 0.4471,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.10923303663730621,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 3556.0,
|
| 494 |
+
"valid_targets_min": 516
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4430176565008026,
|
| 498 |
+
"grad_norm": 0.367939390471383,
|
| 499 |
+
"learning_rate": 3.5880678527966224e-05,
|
| 500 |
+
"loss": 0.4404,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.12570562958717346,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 3408.8,
|
| 505 |
+
"valid_targets_min": 960
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4751203852327448,
|
| 509 |
+
"grad_norm": 0.294671299662254,
|
| 510 |
+
"learning_rate": 3.5604690057839456e-05,
|
| 511 |
+
"loss": 0.4293,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.10687486082315445,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 4010.9,
|
| 516 |
+
"valid_targets_min": 853
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.507223113964687,
|
| 520 |
+
"grad_norm": 0.2958977994024142,
|
| 521 |
+
"learning_rate": 3.532088886237956e-05,
|
| 522 |
+
"loss": 0.4374,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.1045038253068924,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 3799.8,
|
| 527 |
+
"valid_targets_min": 819
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5393258426966292,
|
| 531 |
+
"grad_norm": 0.29728785349279935,
|
| 532 |
+
"learning_rate": 3.5029417030970005e-05,
|
| 533 |
+
"loss": 0.4361,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.10812290012836456,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 3272.2,
|
| 538 |
+
"valid_targets_min": 945
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.5714285714285714,
|
| 542 |
+
"grad_norm": 0.31688903277050234,
|
| 543 |
+
"learning_rate": 3.473042049341474e-05,
|
| 544 |
+
"loss": 0.4323,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.10774408280849457,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 3913.6,
|
| 549 |
+
"valid_targets_min": 1259
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.6035313001605136,
|
| 553 |
+
"grad_norm": 0.3692867105165629,
|
| 554 |
+
"learning_rate": 3.442404894687629e-05,
|
| 555 |
+
"loss": 0.438,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.1150355115532875,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 3968.7,
|
| 560 |
+
"valid_targets_min": 1449
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.635634028892456,
|
| 564 |
+
"grad_norm": 0.29717175235926496,
|
| 565 |
+
"learning_rate": 3.411045578092754e-05,
|
| 566 |
+
"loss": 0.4232,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.11245591938495636,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 3593.1,
|
| 571 |
+
"valid_targets_min": 836
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.667736757624398,
|
| 575 |
+
"grad_norm": 0.3221818053403484,
|
| 576 |
+
"learning_rate": 3.3789798000754924e-05,
|
| 577 |
+
"loss": 0.4362,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.11561841517686844,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 3729.6,
|
| 582 |
+
"valid_targets_min": 1041
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.6998394863563404,
|
| 586 |
+
"grad_norm": 0.32510941540680377,
|
| 587 |
+
"learning_rate": 3.34622361485514e-05,
|
| 588 |
+
"loss": 0.434,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.1104997992515564,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 3289.7,
|
| 593 |
+
"valid_targets_min": 754
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7319422150882824,
|
| 597 |
+
"grad_norm": 0.2850424295372404,
|
| 598 |
+
"learning_rate": 3.31279342231386e-05,
|
| 599 |
+
"loss": 0.4275,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.09886403381824493,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 3479.6,
|
| 604 |
+
"valid_targets_min": 1010
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7640449438202248,
|
| 608 |
+
"grad_norm": 0.29747010204623664,
|
| 609 |
+
"learning_rate": 3.278705959785821e-05,
|
| 610 |
+
"loss": 0.4151,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.09024910628795624,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 2982.7,
|
| 615 |
+
"valid_targets_min": 1051
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7961476725521668,
|
| 619 |
+
"grad_norm": 0.2594879128158275,
|
| 620 |
+
"learning_rate": 3.243978293677408e-05,
|
| 621 |
+
"loss": 0.4082,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.10707713663578033,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 4601.4,
|
| 626 |
+
"valid_targets_min": 1253
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8282504012841092,
|
| 630 |
+
"grad_norm": 0.3092407046323038,
|
| 631 |
+
"learning_rate": 3.208627810922665e-05,
|
| 632 |
+
"loss": 0.436,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.1255902498960495,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 3863.8,
|
| 637 |
+
"valid_targets_min": 1065
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.8603531300160514,
|
| 641 |
+
"grad_norm": 0.33530232091926576,
|
| 642 |
+
"learning_rate": 3.172672210278254e-05,
|
| 643 |
+
"loss": 0.4209,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.08784206211566925,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 3285.4,
|
| 648 |
+
"valid_targets_min": 691
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.8924558587479936,
|
| 652 |
+
"grad_norm": 0.2968668400465552,
|
| 653 |
+
"learning_rate": 3.136129493462312e-05,
|
| 654 |
+
"loss": 0.4319,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.12176284193992615,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 4110.1,
|
| 659 |
+
"valid_targets_min": 753
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.9245585874799358,
|
| 663 |
+
"grad_norm": 0.3002839956959657,
|
| 664 |
+
"learning_rate": 3.0990179561416124e-05,
|
| 665 |
+
"loss": 0.4201,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.10375160723924637,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 3537.5,
|
| 670 |
+
"valid_targets_min": 776
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.956661316211878,
|
| 674 |
+
"grad_norm": 0.29150855964222466,
|
| 675 |
+
"learning_rate": 3.061356178771564e-05,
|
| 676 |
+
"loss": 0.4251,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.08603081107139587,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 2934.9,
|
| 681 |
+
"valid_targets_min": 517
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9887640449438202,
|
| 685 |
+
"grad_norm": 0.31045167265910495,
|
| 686 |
+
"learning_rate": 3.0231630172936205e-05,
|
| 687 |
+
"loss": 0.4147,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.09072297811508179,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 3133.2,
|
| 692 |
+
"valid_targets_min": 983
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.019261637239165,
|
| 696 |
+
"grad_norm": 0.28708543588073493,
|
| 697 |
+
"learning_rate": 2.98445759369477e-05,
|
| 698 |
+
"loss": 0.4074,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.09558726847171783,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 4008.1,
|
| 703 |
+
"valid_targets_min": 515
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0513643659711076,
|
| 707 |
+
"grad_norm": 0.3416876858005374,
|
| 708 |
+
"learning_rate": 2.945259286433819e-05,
|
| 709 |
+
"loss": 0.4149,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.11786194890737534,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 3077.6,
|
| 714 |
+
"valid_targets_min": 520
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0834670947030496,
|
| 718 |
+
"grad_norm": 0.3303316995435761,
|
| 719 |
+
"learning_rate": 2.9055877207392752e-05,
|
| 720 |
+
"loss": 0.4184,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.09221881628036499,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 2562.9,
|
| 725 |
+
"valid_targets_min": 1311
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.115569823434992,
|
| 729 |
+
"grad_norm": 0.3074809241219434,
|
| 730 |
+
"learning_rate": 2.8654627587836724e-05,
|
| 731 |
+
"loss": 0.4253,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.12592661380767822,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 4764.7,
|
| 736 |
+
"valid_targets_min": 1340
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.1476725521669344,
|
| 740 |
+
"grad_norm": 0.3043947470357316,
|
| 741 |
+
"learning_rate": 2.8249044897392814e-05,
|
| 742 |
+
"loss": 0.4071,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.11719249188899994,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 4137.8,
|
| 747 |
+
"valid_targets_min": 1536
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.1797752808988764,
|
| 751 |
+
"grad_norm": 0.2890043796848122,
|
| 752 |
+
"learning_rate": 2.7839332197201505e-05,
|
| 753 |
+
"loss": 0.4119,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.10229069739580154,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 4544.4,
|
| 758 |
+
"valid_targets_min": 1321
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.211878009630819,
|
| 762 |
+
"grad_norm": 0.28710289619081225,
|
| 763 |
+
"learning_rate": 2.7425694616155474e-05,
|
| 764 |
+
"loss": 0.4147,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.09542287886142731,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 3431.3,
|
| 769 |
+
"valid_targets_min": 1012
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.243980738362761,
|
| 773 |
+
"grad_norm": 0.33593483468173724,
|
| 774 |
+
"learning_rate": 2.700833924819868e-05,
|
| 775 |
+
"loss": 0.4137,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.10045792162418365,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 3555.9,
|
| 780 |
+
"valid_targets_min": 868
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.276083467094703,
|
| 784 |
+
"grad_norm": 0.3050460009929215,
|
| 785 |
+
"learning_rate": 2.6587475048641596e-05,
|
| 786 |
+
"loss": 0.421,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.1351967751979828,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 4165.8,
|
| 791 |
+
"valid_targets_min": 864
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.308186195826645,
|
| 795 |
+
"grad_norm": 0.3007356034553604,
|
| 796 |
+
"learning_rate": 2.6163312729544616e-05,
|
| 797 |
+
"loss": 0.4124,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.10512376576662064,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 3216.5,
|
| 802 |
+
"valid_targets_min": 524
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3402889245585876,
|
| 806 |
+
"grad_norm": 0.27621857982103765,
|
| 807 |
+
"learning_rate": 2.5736064654221808e-05,
|
| 808 |
+
"loss": 0.3984,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.12070536613464355,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 4767.9,
|
| 813 |
+
"valid_targets_min": 978
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3723916532905296,
|
| 817 |
+
"grad_norm": 0.28304064443701404,
|
| 818 |
+
"learning_rate": 2.5305944730918017e-05,
|
| 819 |
+
"loss": 0.4133,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.09823143482208252,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 4099.4,
|
| 824 |
+
"valid_targets_min": 1167
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.404494382022472,
|
| 828 |
+
"grad_norm": 0.3317797648395537,
|
| 829 |
+
"learning_rate": 2.487316830571244e-05,
|
| 830 |
+
"loss": 0.4177,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.08248773217201233,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 3434.5,
|
| 835 |
+
"valid_targets_min": 945
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.436597110754414,
|
| 839 |
+
"grad_norm": 0.34105769887242393,
|
| 840 |
+
"learning_rate": 2.4437952054702354e-05,
|
| 841 |
+
"loss": 0.4131,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.08816108107566833,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 3624.6,
|
| 846 |
+
"valid_targets_min": 551
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.4686998394863564,
|
| 850 |
+
"grad_norm": 0.29849113343556627,
|
| 851 |
+
"learning_rate": 2.4000513875520892e-05,
|
| 852 |
+
"loss": 0.4183,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.10371711850166321,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 4422.7,
|
| 857 |
+
"valid_targets_min": 1146
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.5008025682182984,
|
| 861 |
+
"grad_norm": 0.29651959052335053,
|
| 862 |
+
"learning_rate": 2.3561072778243335e-05,
|
| 863 |
+
"loss": 0.4037,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.10162942856550217,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 3326.1,
|
| 868 |
+
"valid_targets_min": 1415
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.532905296950241,
|
| 872 |
+
"grad_norm": 0.30954429361691455,
|
| 873 |
+
"learning_rate": 2.311984877573636e-05,
|
| 874 |
+
"loss": 0.4075,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.09746764600276947,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 3306.4,
|
| 879 |
+
"valid_targets_min": 1062
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5650080256821832,
|
| 883 |
+
"grad_norm": 0.2904497497376163,
|
| 884 |
+
"learning_rate": 2.2677062773505235e-05,
|
| 885 |
+
"loss": 0.4158,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.1188712939620018,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 4573.7,
|
| 890 |
+
"valid_targets_min": 972
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.597110754414125,
|
| 894 |
+
"grad_norm": 0.3164646902586644,
|
| 895 |
+
"learning_rate": 2.2232936459094158e-05,
|
| 896 |
+
"loss": 0.4183,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.1208307221531868,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 3505.8,
|
| 901 |
+
"valid_targets_min": 1253
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.629213483146067,
|
| 905 |
+
"grad_norm": 0.36094852940534805,
|
| 906 |
+
"learning_rate": 2.1787692191094986e-05,
|
| 907 |
+
"loss": 0.3936,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.09573773294687271,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 3226.8,
|
| 912 |
+
"valid_targets_min": 808
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6613162118780096,
|
| 916 |
+
"grad_norm": 0.3494215415068745,
|
| 917 |
+
"learning_rate": 2.1341552887820048e-05,
|
| 918 |
+
"loss": 0.415,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.10980828106403351,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 3792.3,
|
| 923 |
+
"valid_targets_min": 1168
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.693418940609952,
|
| 927 |
+
"grad_norm": 0.3890932782850568,
|
| 928 |
+
"learning_rate": 2.0894741915694685e-05,
|
| 929 |
+
"loss": 0.4111,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.09624197334051132,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 3245.2,
|
| 934 |
+
"valid_targets_min": 1020
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.725521669341894,
|
| 938 |
+
"grad_norm": 0.2721958461218534,
|
| 939 |
+
"learning_rate": 2.0447482977425465e-05,
|
| 940 |
+
"loss": 0.4047,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.0908433347940445,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 3821.1,
|
| 945 |
+
"valid_targets_min": 1278
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7576243980738364,
|
| 949 |
+
"grad_norm": 0.2925048879553211,
|
| 950 |
+
"learning_rate": 2e-05,
|
| 951 |
+
"loss": 0.4148,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.12087635695934296,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 3961.0,
|
| 956 |
+
"valid_targets_min": 848
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.7897271268057784,
|
| 960 |
+
"grad_norm": 0.2931211576473339,
|
| 961 |
+
"learning_rate": 1.9552517022574542e-05,
|
| 962 |
+
"loss": 0.4073,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.10215561091899872,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 4705.2,
|
| 967 |
+
"valid_targets_min": 1519
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.821829855537721,
|
| 971 |
+
"grad_norm": 0.31375433571050343,
|
| 972 |
+
"learning_rate": 1.910525808430532e-05,
|
| 973 |
+
"loss": 0.4198,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.0890008732676506,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 3066.1,
|
| 978 |
+
"valid_targets_min": 1310
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.853932584269663,
|
| 982 |
+
"grad_norm": 0.32732471726116175,
|
| 983 |
+
"learning_rate": 1.8658447112179952e-05,
|
| 984 |
+
"loss": 0.4064,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.1120062842965126,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 3805.1,
|
| 989 |
+
"valid_targets_min": 816
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.886035313001605,
|
| 993 |
+
"grad_norm": 0.32531970253612086,
|
| 994 |
+
"learning_rate": 1.821230780890502e-05,
|
| 995 |
+
"loss": 0.4039,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.08783616870641708,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 3062.5,
|
| 1000 |
+
"valid_targets_min": 930
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.918138041733547,
|
| 1004 |
+
"grad_norm": 0.30368172861725967,
|
| 1005 |
+
"learning_rate": 1.776706354090585e-05,
|
| 1006 |
+
"loss": 0.3946,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.10530835390090942,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 3589.1,
|
| 1011 |
+
"valid_targets_min": 1070
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9502407704654896,
|
| 1015 |
+
"grad_norm": 0.2990061639291643,
|
| 1016 |
+
"learning_rate": 1.7322937226494775e-05,
|
| 1017 |
+
"loss": 0.4167,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.10799364745616913,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 3779.5,
|
| 1022 |
+
"valid_targets_min": 1047
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.982343499197432,
|
| 1026 |
+
"grad_norm": 0.3067377142223239,
|
| 1027 |
+
"learning_rate": 1.6880151224263646e-05,
|
| 1028 |
+
"loss": 0.4143,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.09094572067260742,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 3441.7,
|
| 1033 |
+
"valid_targets_min": 730
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 3.012841091492777,
|
| 1037 |
+
"grad_norm": 0.30726836931947277,
|
| 1038 |
+
"learning_rate": 1.643892722175667e-05,
|
| 1039 |
+
"loss": 0.3996,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.09883762896060944,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 3639.9,
|
| 1044 |
+
"valid_targets_min": 696
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.044943820224719,
|
| 1048 |
+
"grad_norm": 0.2852457832390904,
|
| 1049 |
+
"learning_rate": 1.5999486124479115e-05,
|
| 1050 |
+
"loss": 0.3963,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.12268540263175964,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 5229.9,
|
| 1055 |
+
"valid_targets_min": 1711
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.077046548956661,
|
| 1059 |
+
"grad_norm": 0.34155419020286215,
|
| 1060 |
+
"learning_rate": 1.5562047945297652e-05,
|
| 1061 |
+
"loss": 0.3921,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.09211775660514832,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 2812.4,
|
| 1066 |
+
"valid_targets_min": 1036
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.1091492776886036,
|
| 1070 |
+
"grad_norm": 0.2937900120598192,
|
| 1071 |
+
"learning_rate": 1.5126831694287564e-05,
|
| 1072 |
+
"loss": 0.3875,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.11049457639455795,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 3831.8,
|
| 1077 |
+
"valid_targets_min": 1421
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.1412520064205456,
|
| 1081 |
+
"grad_norm": 0.3004782410044986,
|
| 1082 |
+
"learning_rate": 1.4694055269081987e-05,
|
| 1083 |
+
"loss": 0.3897,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.09211912006139755,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 3683.4,
|
| 1088 |
+
"valid_targets_min": 896
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.173354735152488,
|
| 1092 |
+
"grad_norm": 0.28620629746496773,
|
| 1093 |
+
"learning_rate": 1.4263935345778202e-05,
|
| 1094 |
+
"loss": 0.4068,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.1081249862909317,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 4564.4,
|
| 1099 |
+
"valid_targets_min": 1295
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.20545746388443,
|
| 1103 |
+
"grad_norm": 0.3065292114147679,
|
| 1104 |
+
"learning_rate": 1.3836687270455388e-05,
|
| 1105 |
+
"loss": 0.4003,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.1026119515299797,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 3764.5,
|
| 1110 |
+
"valid_targets_min": 1524
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.2375601926163724,
|
| 1114 |
+
"grad_norm": 0.2862455916184854,
|
| 1115 |
+
"learning_rate": 1.341252495135841e-05,
|
| 1116 |
+
"loss": 0.404,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.09894335269927979,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 4383.1,
|
| 1121 |
+
"valid_targets_min": 1030
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2696629213483144,
|
| 1125 |
+
"grad_norm": 0.3238338379491629,
|
| 1126 |
+
"learning_rate": 1.299166075180133e-05,
|
| 1127 |
+
"loss": 0.4105,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.11098513007164001,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 3665.1,
|
| 1132 |
+
"valid_targets_min": 1206
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.301765650080257,
|
| 1136 |
+
"grad_norm": 0.31425131498588266,
|
| 1137 |
+
"learning_rate": 1.2574305383844528e-05,
|
| 1138 |
+
"loss": 0.4021,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.08589449524879456,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 3058.8,
|
| 1143 |
+
"valid_targets_min": 773
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.333868378812199,
|
| 1147 |
+
"grad_norm": 0.2970738358435746,
|
| 1148 |
+
"learning_rate": 1.2160667802798503e-05,
|
| 1149 |
+
"loss": 0.4041,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.09149523079395294,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 3748.1,
|
| 1154 |
+
"valid_targets_min": 1046
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.365971107544141,
|
| 1158 |
+
"grad_norm": 0.2780590370917071,
|
| 1159 |
+
"learning_rate": 1.1750955102607193e-05,
|
| 1160 |
+
"loss": 0.4017,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.09089787304401398,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 3762.3,
|
| 1165 |
+
"valid_targets_min": 888
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.3980738362760836,
|
| 1169 |
+
"grad_norm": 0.30125085043484123,
|
| 1170 |
+
"learning_rate": 1.1345372412163276e-05,
|
| 1171 |
+
"loss": 0.4022,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.09668652713298798,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 3652.8,
|
| 1176 |
+
"valid_targets_min": 1198
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4301765650080256,
|
| 1180 |
+
"grad_norm": 0.3393832899109687,
|
| 1181 |
+
"learning_rate": 1.094412279260726e-05,
|
| 1182 |
+
"loss": 0.4071,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.09544992446899414,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 3167.6,
|
| 1187 |
+
"valid_targets_min": 1258
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.462279293739968,
|
| 1191 |
+
"grad_norm": 0.2950846626090308,
|
| 1192 |
+
"learning_rate": 1.0547407135661816e-05,
|
| 1193 |
+
"loss": 0.4,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.1026640385389328,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 4445.8,
|
| 1198 |
+
"valid_targets_min": 1229
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.49438202247191,
|
| 1202 |
+
"grad_norm": 0.2841748611048485,
|
| 1203 |
+
"learning_rate": 1.0155424063052306e-05,
|
| 1204 |
+
"loss": 0.3857,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.09762070327997208,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 4222.4,
|
| 1209 |
+
"valid_targets_min": 1269
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.5264847512038524,
|
| 1213 |
+
"grad_norm": 0.3352654173394246,
|
| 1214 |
+
"learning_rate": 9.768369827063803e-06,
|
| 1215 |
+
"loss": 0.4029,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.11695688217878342,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 3240.5,
|
| 1220 |
+
"valid_targets_min": 932
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5585874799357944,
|
| 1224 |
+
"grad_norm": 0.2806686898678227,
|
| 1225 |
+
"learning_rate": 9.386438212284372e-06,
|
| 1226 |
+
"loss": 0.4033,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.1005910113453865,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 3759.2,
|
| 1231 |
+
"valid_targets_min": 668
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.590690208667737,
|
| 1235 |
+
"grad_norm": 0.33872852815646304,
|
| 1236 |
+
"learning_rate": 9.009820438583881e-06,
|
| 1237 |
+
"loss": 0.4145,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.12395333498716354,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 3597.6,
|
| 1242 |
+
"valid_targets_min": 1306
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6227929373996792,
|
| 1246 |
+
"grad_norm": 0.2654332382202926,
|
| 1247 |
+
"learning_rate": 8.638705065376887e-06,
|
| 1248 |
+
"loss": 0.3852,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.08896806091070175,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 3946.9,
|
| 1253 |
+
"valid_targets_min": 1110
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.654895666131621,
|
| 1257 |
+
"grad_norm": 0.27981716529155093,
|
| 1258 |
+
"learning_rate": 8.27327789721746e-06,
|
| 1259 |
+
"loss": 0.376,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.08400297164916992,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 4350.8,
|
| 1264 |
+
"valid_targets_min": 931
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.686998394863563,
|
| 1268 |
+
"grad_norm": 0.3385370692244635,
|
| 1269 |
+
"learning_rate": 7.913721890773354e-06,
|
| 1270 |
+
"loss": 0.4033,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.12360556423664093,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 3591.9,
|
| 1275 |
+
"valid_targets_min": 1549
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.7191011235955056,
|
| 1279 |
+
"grad_norm": 0.2655051672149985,
|
| 1280 |
+
"learning_rate": 7.560217063225923e-06,
|
| 1281 |
+
"loss": 0.398,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.08802102506160736,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 4639.4,
|
| 1286 |
+
"valid_targets_min": 1643
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.751203852327448,
|
| 1290 |
+
"grad_norm": 0.40923046291632936,
|
| 1291 |
+
"learning_rate": 7.212940402141808e-06,
|
| 1292 |
+
"loss": 0.4014,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.09749025106430054,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 3517.1,
|
| 1297 |
+
"valid_targets_min": 1306
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.78330658105939,
|
| 1301 |
+
"grad_norm": 0.3117388859264426,
|
| 1302 |
+
"learning_rate": 6.8720657768614115e-06,
|
| 1303 |
+
"loss": 0.4097,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.11236117780208588,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 3592.8,
|
| 1308 |
+
"valid_targets_min": 529
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.8154093097913324,
|
| 1312 |
+
"grad_norm": 0.31179165172571244,
|
| 1313 |
+
"learning_rate": 6.537763851448593e-06,
|
| 1314 |
+
"loss": 0.4046,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.10635977983474731,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 3038.1,
|
| 1319 |
+
"valid_targets_min": 1038
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.8475120385232744,
|
| 1323 |
+
"grad_norm": 0.30097213091658315,
|
| 1324 |
+
"learning_rate": 6.210201999245087e-06,
|
| 1325 |
+
"loss": 0.4153,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.09912356734275818,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 3369.2,
|
| 1330 |
+
"valid_targets_min": 301
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.879614767255217,
|
| 1334 |
+
"grad_norm": 0.3092144816503789,
|
| 1335 |
+
"learning_rate": 5.889544219072465e-06,
|
| 1336 |
+
"loss": 0.4028,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.10438457131385803,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 3691.4,
|
| 1341 |
+
"valid_targets_min": 1000
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.911717495987159,
|
| 1345 |
+
"grad_norm": 0.28578697194711516,
|
| 1346 |
+
"learning_rate": 5.5759510531237116e-06,
|
| 1347 |
+
"loss": 0.3933,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.09700490534305573,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 3770.8,
|
| 1352 |
+
"valid_targets_min": 1355
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.943820224719101,
|
| 1356 |
+
"grad_norm": 0.3272428302409307,
|
| 1357 |
+
"learning_rate": 5.269579506585259e-06,
|
| 1358 |
+
"loss": 0.3911,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.08706691116094589,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 3215.6,
|
| 1363 |
+
"valid_targets_min": 837
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.975922953451043,
|
| 1367 |
+
"grad_norm": 0.2811671525014539,
|
| 1368 |
+
"learning_rate": 4.97058296903e-06,
|
| 1369 |
+
"loss": 0.401,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.10784287750720978,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 3762.6,
|
| 1374 |
+
"valid_targets_min": 845
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.006420545746389,
|
| 1378 |
+
"grad_norm": 0.2917034910670569,
|
| 1379 |
+
"learning_rate": 4.679111137620442e-06,
|
| 1380 |
+
"loss": 0.3997,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.09262470155954361,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 3377.3,
|
| 1385 |
+
"valid_targets_min": 896
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.03852327447833,
|
| 1389 |
+
"grad_norm": 0.31513354693746726,
|
| 1390 |
+
"learning_rate": 4.395309942160551e-06,
|
| 1391 |
+
"loss": 0.3913,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.10473042726516724,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 3944.6,
|
| 1396 |
+
"valid_targets_min": 551
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.070626003210273,
|
| 1400 |
+
"grad_norm": 0.27934962375420436,
|
| 1401 |
+
"learning_rate": 4.119321472033779e-06,
|
| 1402 |
+
"loss": 0.3949,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.0756620317697525,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 2551.7,
|
| 1407 |
+
"valid_targets_min": 536
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.102728731942215,
|
| 1411 |
+
"grad_norm": 0.37095369408284645,
|
| 1412 |
+
"learning_rate": 3.851283905063863e-06,
|
| 1413 |
+
"loss": 0.3946,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.12161412835121155,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 3931.2,
|
| 1418 |
+
"valid_targets_min": 1237
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.134831460674158,
|
| 1422 |
+
"grad_norm": 0.27970269585114166,
|
| 1423 |
+
"learning_rate": 3.5913314383339937e-06,
|
| 1424 |
+
"loss": 0.3872,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.09911416471004486,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 4231.0,
|
| 1429 |
+
"valid_targets_min": 1253
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.166934189406099,
|
| 1433 |
+
"grad_norm": 0.3456613071245264,
|
| 1434 |
+
"learning_rate": 3.3395942209989694e-06,
|
| 1435 |
+
"loss": 0.4023,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.08239921182394028,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 2619.8,
|
| 1440 |
+
"valid_targets_min": 886
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.199036918138042,
|
| 1444 |
+
"grad_norm": 0.2768716249831359,
|
| 1445 |
+
"learning_rate": 3.0961982891241083e-06,
|
| 1446 |
+
"loss": 0.3731,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.09643545746803284,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 3902.5,
|
| 1451 |
+
"valid_targets_min": 1105
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.231139646869984,
|
| 1455 |
+
"grad_norm": 0.3014051499821219,
|
| 1456 |
+
"learning_rate": 2.8612655025833148e-06,
|
| 1457 |
+
"loss": 0.395,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.11026032269001007,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 4226.5,
|
| 1462 |
+
"valid_targets_min": 982
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.263242375601926,
|
| 1466 |
+
"grad_norm": 0.28379604796854446,
|
| 1467 |
+
"learning_rate": 2.6349134840481294e-06,
|
| 1468 |
+
"loss": 0.3809,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.10869620740413666,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 4199.8,
|
| 1473 |
+
"valid_targets_min": 1291
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.295345104333869,
|
| 1477 |
+
"grad_norm": 0.32690267770951353,
|
| 1478 |
+
"learning_rate": 2.417255560098111e-06,
|
| 1479 |
+
"loss": 0.3984,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.08836868405342102,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 3305.6,
|
| 1484 |
+
"valid_targets_min": 561
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.32744783306581,
|
| 1488 |
+
"grad_norm": 0.25345525186658413,
|
| 1489 |
+
"learning_rate": 2.2084007044821764e-06,
|
| 1490 |
+
"loss": 0.389,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.09824281930923462,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 4750.1,
|
| 1495 |
+
"valid_targets_min": 1209
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.359550561797753,
|
| 1499 |
+
"grad_norm": 0.2689895453757197,
|
| 1500 |
+
"learning_rate": 2.0084534835592584e-06,
|
| 1501 |
+
"loss": 0.3961,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.10501833260059357,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 4509.6,
|
| 1506 |
+
"valid_targets_min": 1157
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.391653290529695,
|
| 1510 |
+
"grad_norm": 0.27247257469639,
|
| 1511 |
+
"learning_rate": 1.817514003945524e-06,
|
| 1512 |
+
"loss": 0.4008,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.09853364527225494,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 4158.9,
|
| 1517 |
+
"valid_targets_min": 726
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.423756019261638,
|
| 1521 |
+
"grad_norm": 0.31501055011520845,
|
| 1522 |
+
"learning_rate": 1.6356778623945223e-06,
|
| 1523 |
+
"loss": 0.4052,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.1491430252790451,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 4383.5,
|
| 1528 |
+
"valid_targets_min": 1456
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.455858747993579,
|
| 1532 |
+
"grad_norm": 0.2715083485471513,
|
| 1533 |
+
"learning_rate": 1.4630360979351644e-06,
|
| 1534 |
+
"loss": 0.4089,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.10639771819114685,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 4281.1,
|
| 1539 |
+
"valid_targets_min": 1882
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.487961476725522,
|
| 1543 |
+
"grad_norm": 0.269592691484671,
|
| 1544 |
+
"learning_rate": 1.2996751462917057e-06,
|
| 1545 |
+
"loss": 0.4004,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.095083087682724,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 3574.7,
|
| 1550 |
+
"valid_targets_min": 934
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.520064205457464,
|
| 1554 |
+
"grad_norm": 0.28970840272710097,
|
| 1555 |
+
"learning_rate": 1.1456767966083393e-06,
|
| 1556 |
+
"loss": 0.3842,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.10370156168937683,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 3411.7,
|
| 1561 |
+
"valid_targets_min": 529
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.552166934189406,
|
| 1565 |
+
"grad_norm": 0.2870909780240681,
|
| 1566 |
+
"learning_rate": 1.0011181505002488e-06,
|
| 1567 |
+
"loss": 0.3908,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.10310973972082138,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 4500.2,
|
| 1572 |
+
"valid_targets_min": 1171
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.584269662921348,
|
| 1576 |
+
"grad_norm": 0.2688461264330106,
|
| 1577 |
+
"learning_rate": 8.660715834514977e-07,
|
| 1578 |
+
"loss": 0.3843,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.08639390021562576,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 3641.4,
|
| 1583 |
+
"valid_targets_min": 922
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.61637239165329,
|
| 1587 |
+
"grad_norm": 0.25765194882082687,
|
| 1588 |
+
"learning_rate": 7.406047085791379e-07,
|
| 1589 |
+
"loss": 0.3861,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.0939197838306427,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 4168.2,
|
| 1594 |
+
"valid_targets_min": 1460
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.648475120385233,
|
| 1598 |
+
"grad_norm": 0.30160125278292577,
|
| 1599 |
+
"learning_rate": 6.247803427816945e-07,
|
| 1600 |
+
"loss": 0.4133,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.11861149966716766,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 3679.9,
|
| 1605 |
+
"valid_targets_min": 1359
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.680577849117175,
|
| 1609 |
+
"grad_norm": 0.2793314638141966,
|
| 1610 |
+
"learning_rate": 5.186564752889078e-07,
|
| 1611 |
+
"loss": 0.3965,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.09253215789794922,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 3437.8,
|
| 1616 |
+
"valid_targets_min": 1053
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.712680577849117,
|
| 1620 |
+
"grad_norm": 0.27123116724226537,
|
| 1621 |
+
"learning_rate": 4.22286238628562e-07,
|
| 1622 |
+
"loss": 0.3784,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.07941226661205292,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 3344.3,
|
| 1627 |
+
"valid_targets_min": 342
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.744783306581059,
|
| 1631 |
+
"grad_norm": 0.2785272531826042,
|
| 1632 |
+
"learning_rate": 3.357178820248641e-07,
|
| 1633 |
+
"loss": 0.3961,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.08652195334434509,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 3726.1,
|
| 1638 |
+
"valid_targets_min": 1384
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.776886035313002,
|
| 1642 |
+
"grad_norm": 0.31544979123824807,
|
| 1643 |
+
"learning_rate": 2.5899474724174313e-07,
|
| 1644 |
+
"loss": 0.3988,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.08774863183498383,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 2830.6,
|
| 1649 |
+
"valid_targets_min": 467
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.808988764044944,
|
| 1653 |
+
"grad_norm": 0.296446741850878,
|
| 1654 |
+
"learning_rate": 1.9215524688313893e-07,
|
| 1655 |
+
"loss": 0.3927,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.10991404950618744,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 3905.6,
|
| 1660 |
+
"valid_targets_min": 516
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.841091492776886,
|
| 1664 |
+
"grad_norm": 0.27186316337463917,
|
| 1665 |
+
"learning_rate": 1.3523284516113955e-07,
|
| 1666 |
+
"loss": 0.387,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.09752750396728516,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 4108.7,
|
| 1671 |
+
"valid_targets_min": 962
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.873194221508828,
|
| 1675 |
+
"grad_norm": 0.3001939708765642,
|
| 1676 |
+
"learning_rate": 8.825604114162556e-08,
|
| 1677 |
+
"loss": 0.3918,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.08993087708950043,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 3836.1,
|
| 1682 |
+
"valid_targets_min": 1251
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.90529695024077,
|
| 1686 |
+
"grad_norm": 0.2867156389445644,
|
| 1687 |
+
"learning_rate": 5.1248354475768034e-08,
|
| 1688 |
+
"loss": 0.4068,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.09052161872386932,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 3208.5,
|
| 1693 |
+
"valid_targets_min": 1013
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.937399678972713,
|
| 1697 |
+
"grad_norm": 0.2766893813163677,
|
| 1698 |
+
"learning_rate": 2.4228313624568013e-08,
|
| 1699 |
+
"loss": 0.3876,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.1123722642660141,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 5074.8,
|
| 1704 |
+
"valid_targets_min": 1145
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.969502407704655,
|
| 1708 |
+
"grad_norm": 0.339877298742656,
|
| 1709 |
+
"learning_rate": 7.209446582292501e-09,
|
| 1710 |
+
"loss": 0.4201,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.11707846820354462,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 3143.6,
|
| 1715 |
+
"valid_targets_min": 998
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 5.0,
|
| 1719 |
+
"grad_norm": 0.34958645297704605,
|
| 1720 |
+
"learning_rate": 2.002741034812239e-10,
|
| 1721 |
+
"loss": 0.4005,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.11478559672832489,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 3291.8,
|
| 1726 |
+
"valid_targets_min": 500
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"loss_nan_ranks": 0,
|
| 1731 |
+
"loss_rank_avg": 0.11478559672832489,
|
| 1732 |
+
"step": 780,
|
| 1733 |
+
"total_flos": 1.0263973305862062e+18,
|
| 1734 |
+
"train_loss": 0.4329962073228298,
|
| 1735 |
+
"train_runtime": 13926.0897,
|
| 1736 |
+
"train_samples_per_second": 3.575,
|
| 1737 |
+
"train_steps_per_second": 0.056,
|
| 1738 |
+
"valid_targets_mean": 3291.8,
|
| 1739 |
+
"valid_targets_min": 500
|
| 1740 |
+
}
|
| 1741 |
+
],
|
| 1742 |
+
"logging_steps": 5,
|
| 1743 |
+
"max_steps": 780,
|
| 1744 |
+
"num_input_tokens_seen": 0,
|
| 1745 |
+
"num_train_epochs": 5,
|
| 1746 |
+
"save_steps": 500,
|
| 1747 |
+
"stateful_callbacks": {
|
| 1748 |
+
"TrainerControl": {
|
| 1749 |
+
"args": {
|
| 1750 |
+
"should_epoch_stop": false,
|
| 1751 |
+
"should_evaluate": false,
|
| 1752 |
+
"should_log": false,
|
| 1753 |
+
"should_save": false,
|
| 1754 |
+
"should_training_stop": false
|
| 1755 |
+
},
|
| 1756 |
+
"attributes": {}
|
| 1757 |
+
}
|
| 1758 |
+
},
|
| 1759 |
+
"total_flos": 1.0263973305862062e+18,
|
| 1760 |
+
"train_batch_size": 1,
|
| 1761 |
+
"trial_name": null,
|
| 1762 |
+
"trial_params": null
|
| 1763 |
+
}
|
training_loss.png
ADDED
|