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: code-contests-sandboxes-traces-terminus-2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# code-contests-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: code-contests-sandboxes-traces-terminus-2
|
|
|
|
| 16 |
|
| 17 |
# code-contests-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/code-contests-sandboxes-traces-terminus-2 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.010832708519495,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 155.84360988986572,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.36648064851760864,
|
| 7 |
+
"mfu_percent": 0.28345107480703147,
|
| 8 |
+
"mfu_percent_theoretical": 11.013682677729026,
|
| 9 |
+
"total_flos": 1.0910552897523548e+18,
|
| 10 |
+
"train_loss": 0.3873556343613157,
|
| 11 |
+
"train_runtime": 17001.6953,
|
| 12 |
+
"train_samples_per_second": 2.941,
|
| 13 |
+
"train_steps_per_second": 0.046,
|
| 14 |
+
"valid_targets_mean": 3627.8,
|
| 15 |
+
"valid_targets_min": 954
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.010832708519495,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 155.84360988986572,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.36648064851760864,
|
| 7 |
+
"mfu_percent": 0.28345107480703147,
|
| 8 |
+
"mfu_percent_theoretical": 11.013682677729026,
|
| 9 |
+
"total_flos": 1.0910552897523548e+18,
|
| 10 |
+
"train_loss": 0.3873556343613157,
|
| 11 |
+
"train_runtime": 17001.6953,
|
| 12 |
+
"train_samples_per_second": 2.941,
|
| 13 |
+
"train_steps_per_second": 0.046,
|
| 14 |
+
"valid_targets_mean": 3627.8,
|
| 15 |
+
"valid_targets_min": 954
|
| 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": 5.442185453929223,
|
| 15 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 16 |
+
"loss": 0.6905,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.19166269898414612,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 3772.9,
|
| 21 |
+
"valid_targets_min": 1172
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.064,
|
| 25 |
+
"grad_norm": 3.8095328022653705,
|
| 26 |
+
"learning_rate": 4.556962025316456e-06,
|
| 27 |
+
"loss": 0.6771,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.20748385787010193,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 5537.4,
|
| 32 |
+
"valid_targets_min": 684
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.096,
|
| 36 |
+
"grad_norm": 2.1200115743319388,
|
| 37 |
+
"learning_rate": 7.08860759493671e-06,
|
| 38 |
+
"loss": 0.639,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.16293901205062866,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 4067.4,
|
| 43 |
+
"valid_targets_min": 422
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.128,
|
| 47 |
+
"grad_norm": 0.9257072627481844,
|
| 48 |
+
"learning_rate": 9.620253164556963e-06,
|
| 49 |
+
"loss": 0.5846,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.1487666368484497,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 3707.1,
|
| 54 |
+
"valid_targets_min": 905
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.16,
|
| 58 |
+
"grad_norm": 0.7988850073269412,
|
| 59 |
+
"learning_rate": 1.2151898734177216e-05,
|
| 60 |
+
"loss": 0.5754,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.19959388673305511,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 5725.4,
|
| 65 |
+
"valid_targets_min": 665
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.192,
|
| 69 |
+
"grad_norm": 0.5331769320628815,
|
| 70 |
+
"learning_rate": 1.468354430379747e-05,
|
| 71 |
+
"loss": 0.5604,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.12151822447776794,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 3568.0,
|
| 76 |
+
"valid_targets_min": 727
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.224,
|
| 80 |
+
"grad_norm": 0.45282811415327257,
|
| 81 |
+
"learning_rate": 1.7215189873417723e-05,
|
| 82 |
+
"loss": 0.5278,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.11509503424167633,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 3148.8,
|
| 87 |
+
"valid_targets_min": 1224
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.256,
|
| 91 |
+
"grad_norm": 0.4736002691754016,
|
| 92 |
+
"learning_rate": 1.974683544303798e-05,
|
| 93 |
+
"loss": 0.5098,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.12758716940879822,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 3358.9,
|
| 98 |
+
"valid_targets_min": 811
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.288,
|
| 102 |
+
"grad_norm": 0.31632295133430655,
|
| 103 |
+
"learning_rate": 2.2278481012658228e-05,
|
| 104 |
+
"loss": 0.4781,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.08016407489776611,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 3059.4,
|
| 109 |
+
"valid_targets_min": 790
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32,
|
| 113 |
+
"grad_norm": 0.348782345524125,
|
| 114 |
+
"learning_rate": 2.481012658227848e-05,
|
| 115 |
+
"loss": 0.4745,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.13580970466136932,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 4653.7,
|
| 120 |
+
"valid_targets_min": 1002
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.352,
|
| 124 |
+
"grad_norm": 0.25342842121431647,
|
| 125 |
+
"learning_rate": 2.7341772151898737e-05,
|
| 126 |
+
"loss": 0.4667,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.12497734278440475,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 5929.0,
|
| 131 |
+
"valid_targets_min": 737
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.384,
|
| 135 |
+
"grad_norm": 0.2869586553254826,
|
| 136 |
+
"learning_rate": 2.987341772151899e-05,
|
| 137 |
+
"loss": 0.4442,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.1034400537610054,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 2641.9,
|
| 142 |
+
"valid_targets_min": 651
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.416,
|
| 146 |
+
"grad_norm": 0.3279100223436585,
|
| 147 |
+
"learning_rate": 3.240506329113924e-05,
|
| 148 |
+
"loss": 0.4432,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.07635253667831421,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 2183.4,
|
| 153 |
+
"valid_targets_min": 624
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.448,
|
| 157 |
+
"grad_norm": 0.27304560909035697,
|
| 158 |
+
"learning_rate": 3.49367088607595e-05,
|
| 159 |
+
"loss": 0.4394,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.1170448511838913,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 4851.8,
|
| 164 |
+
"valid_targets_min": 920
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48,
|
| 168 |
+
"grad_norm": 0.2666107311165606,
|
| 169 |
+
"learning_rate": 3.746835443037975e-05,
|
| 170 |
+
"loss": 0.4498,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.12639470398426056,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 5269.2,
|
| 175 |
+
"valid_targets_min": 699
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.512,
|
| 179 |
+
"grad_norm": 0.2772064811496187,
|
| 180 |
+
"learning_rate": 4e-05,
|
| 181 |
+
"loss": 0.4298,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.0887710303068161,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 2959.7,
|
| 186 |
+
"valid_targets_min": 586
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.544,
|
| 190 |
+
"grad_norm": 0.2544028411955918,
|
| 191 |
+
"learning_rate": 3.999504991751045e-05,
|
| 192 |
+
"loss": 0.4136,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.12488332390785217,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 5450.7,
|
| 197 |
+
"valid_targets_min": 693
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.576,
|
| 201 |
+
"grad_norm": 0.27618561723531765,
|
| 202 |
+
"learning_rate": 3.9980202120373464e-05,
|
| 203 |
+
"loss": 0.4111,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.07022757828235626,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 2691.6,
|
| 208 |
+
"valid_targets_min": 878
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.608,
|
| 212 |
+
"grad_norm": 0.29279387688561254,
|
| 213 |
+
"learning_rate": 3.995546395837111e-05,
|
| 214 |
+
"loss": 0.4225,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.1592681109905243,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 5913.5,
|
| 219 |
+
"valid_targets_min": 789
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.64,
|
| 223 |
+
"grad_norm": 0.3076526359692288,
|
| 224 |
+
"learning_rate": 3.992084767709763e-05,
|
| 225 |
+
"loss": 0.4071,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.08246097713708878,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 3579.7,
|
| 230 |
+
"valid_targets_min": 598
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.672,
|
| 234 |
+
"grad_norm": 0.3196948304195173,
|
| 235 |
+
"learning_rate": 3.987637041189781e-05,
|
| 236 |
+
"loss": 0.4145,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.07396609336137772,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 2704.1,
|
| 241 |
+
"valid_targets_min": 784
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.704,
|
| 245 |
+
"grad_norm": 0.27897727552461854,
|
| 246 |
+
"learning_rate": 3.982205417938482e-05,
|
| 247 |
+
"loss": 0.4194,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.1036178469657898,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 4052.7,
|
| 252 |
+
"valid_targets_min": 692
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.736,
|
| 256 |
+
"grad_norm": 0.2999073837951396,
|
| 257 |
+
"learning_rate": 3.975792586654179e-05,
|
| 258 |
+
"loss": 0.4139,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.08229044079780579,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 3082.8,
|
| 263 |
+
"valid_targets_min": 623
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.768,
|
| 267 |
+
"grad_norm": 0.3358098936207653,
|
| 268 |
+
"learning_rate": 3.968401721741259e-05,
|
| 269 |
+
"loss": 0.4227,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.14959537982940674,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 4226.0,
|
| 274 |
+
"valid_targets_min": 701
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8,
|
| 278 |
+
"grad_norm": 0.2727171262170047,
|
| 279 |
+
"learning_rate": 3.960036481738819e-05,
|
| 280 |
+
"loss": 0.4236,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.0725155621767044,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 2769.6,
|
| 285 |
+
"valid_targets_min": 632
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.832,
|
| 289 |
+
"grad_norm": 0.2728697189925512,
|
| 290 |
+
"learning_rate": 3.950701007509667e-05,
|
| 291 |
+
"loss": 0.4219,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.09483879804611206,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 3119.8,
|
| 296 |
+
"valid_targets_min": 1146
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.864,
|
| 300 |
+
"grad_norm": 0.27493747655842643,
|
| 301 |
+
"learning_rate": 3.940399920190552e-05,
|
| 302 |
+
"loss": 0.4051,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.07629527151584625,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 4574.4,
|
| 307 |
+
"valid_targets_min": 574
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.896,
|
| 311 |
+
"grad_norm": 0.273673015320205,
|
| 312 |
+
"learning_rate": 3.92913831890467e-05,
|
| 313 |
+
"loss": 0.4214,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.09848488122224808,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 4396.8,
|
| 318 |
+
"valid_targets_min": 779
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.928,
|
| 322 |
+
"grad_norm": 0.2835354881684314,
|
| 323 |
+
"learning_rate": 3.916921778237556e-05,
|
| 324 |
+
"loss": 0.4352,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.06863738596439362,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 2600.9,
|
| 329 |
+
"valid_targets_min": 884
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.96,
|
| 333 |
+
"grad_norm": 0.24549306111748626,
|
| 334 |
+
"learning_rate": 3.903756345477612e-05,
|
| 335 |
+
"loss": 0.415,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.10824453830718994,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 5110.8,
|
| 340 |
+
"valid_targets_min": 725
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.992,
|
| 344 |
+
"grad_norm": 0.26293935756408227,
|
| 345 |
+
"learning_rate": 3.889648537622657e-05,
|
| 346 |
+
"loss": 0.4054,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.08755393326282501,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 4606.3,
|
| 351 |
+
"valid_targets_min": 771
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0192,
|
| 355 |
+
"grad_norm": 0.28805917974478973,
|
| 356 |
+
"learning_rate": 3.874605338153952e-05,
|
| 357 |
+
"loss": 0.3989,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.08316570520401001,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 3181.9,
|
| 362 |
+
"valid_targets_min": 825
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0512,
|
| 366 |
+
"grad_norm": 0.26820639034503513,
|
| 367 |
+
"learning_rate": 3.8586341935793265e-05,
|
| 368 |
+
"loss": 0.3909,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.08125826716423035,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 4078.8,
|
| 373 |
+
"valid_targets_min": 956
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0832,
|
| 377 |
+
"grad_norm": 0.2799230229956331,
|
| 378 |
+
"learning_rate": 3.841743009747089e-05,
|
| 379 |
+
"loss": 0.3972,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.09274695068597794,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 3781.8,
|
| 384 |
+
"valid_targets_min": 983
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1152,
|
| 388 |
+
"grad_norm": 0.26205953485259365,
|
| 389 |
+
"learning_rate": 3.8239401479325714e-05,
|
| 390 |
+
"loss": 0.4088,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.07772032916545868,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 2916.4,
|
| 395 |
+
"valid_targets_min": 571
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1472,
|
| 399 |
+
"grad_norm": 0.2637599374407512,
|
| 400 |
+
"learning_rate": 3.8052344206992276e-05,
|
| 401 |
+
"loss": 0.3943,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.09445404261350632,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 4496.7,
|
| 406 |
+
"valid_targets_min": 800
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1792,
|
| 410 |
+
"grad_norm": 0.2932291816861461,
|
| 411 |
+
"learning_rate": 3.7856350875363396e-05,
|
| 412 |
+
"loss": 0.3781,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.09436005353927612,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 3035.1,
|
| 417 |
+
"valid_targets_min": 693
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2112,
|
| 421 |
+
"grad_norm": 0.27550995248750926,
|
| 422 |
+
"learning_rate": 3.765151850275497e-05,
|
| 423 |
+
"loss": 0.3979,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.10137677937746048,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 4430.1,
|
| 428 |
+
"valid_targets_min": 368
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.2432,
|
| 432 |
+
"grad_norm": 0.2639313244357049,
|
| 433 |
+
"learning_rate": 3.7437948482881104e-05,
|
| 434 |
+
"loss": 0.3969,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.08653558790683746,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 4251.3,
|
| 439 |
+
"valid_targets_min": 563
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2752,
|
| 443 |
+
"grad_norm": 0.3362952524470988,
|
| 444 |
+
"learning_rate": 3.721574653466336e-05,
|
| 445 |
+
"loss": 0.3867,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.11065793037414551,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 4629.8,
|
| 450 |
+
"valid_targets_min": 706
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3072,
|
| 454 |
+
"grad_norm": 0.2650542140015465,
|
| 455 |
+
"learning_rate": 3.698502264989903e-05,
|
| 456 |
+
"loss": 0.3858,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.10903624445199966,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 4045.3,
|
| 461 |
+
"valid_targets_min": 841
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.3392,
|
| 465 |
+
"grad_norm": 0.3676272929156744,
|
| 466 |
+
"learning_rate": 3.674589103881432e-05,
|
| 467 |
+
"loss": 0.3987,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.0916772335767746,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 2358.5,
|
| 472 |
+
"valid_targets_min": 489
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3712,
|
| 476 |
+
"grad_norm": 0.29610462749256095,
|
| 477 |
+
"learning_rate": 3.64984700735293e-05,
|
| 478 |
+
"loss": 0.3786,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.14761430025100708,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 5930.5,
|
| 483 |
+
"valid_targets_min": 512
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4032,
|
| 487 |
+
"grad_norm": 0.2949118655484367,
|
| 488 |
+
"learning_rate": 3.624288222946273e-05,
|
| 489 |
+
"loss": 0.3777,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.10777334123849869,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 4821.2,
|
| 494 |
+
"valid_targets_min": 632
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4352,
|
| 498 |
+
"grad_norm": 0.2600378365295041,
|
| 499 |
+
"learning_rate": 3.597925402470578e-05,
|
| 500 |
+
"loss": 0.3941,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.078704833984375,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 4813.6,
|
| 505 |
+
"valid_targets_min": 695
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4672,
|
| 509 |
+
"grad_norm": 0.24938523801103532,
|
| 510 |
+
"learning_rate": 3.570771595739445e-05,
|
| 511 |
+
"loss": 0.3786,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.08179652690887451,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 4133.4,
|
| 516 |
+
"valid_targets_min": 662
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.4992,
|
| 520 |
+
"grad_norm": 0.29863052946621094,
|
| 521 |
+
"learning_rate": 3.5428402441111964e-05,
|
| 522 |
+
"loss": 0.3749,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.10285639762878418,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 3773.5,
|
| 527 |
+
"valid_targets_min": 661
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5312000000000001,
|
| 531 |
+
"grad_norm": 0.26854744897753935,
|
| 532 |
+
"learning_rate": 3.5141451738352936e-05,
|
| 533 |
+
"loss": 0.3791,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.09314319491386414,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 3503.4,
|
| 538 |
+
"valid_targets_min": 690
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.5632000000000001,
|
| 542 |
+
"grad_norm": 0.354465503459906,
|
| 543 |
+
"learning_rate": 3.4847005892082266e-05,
|
| 544 |
+
"loss": 0.3807,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.07929793000221252,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 2074.6,
|
| 549 |
+
"valid_targets_min": 808
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.5952,
|
| 553 |
+
"grad_norm": 0.2527042035107904,
|
| 554 |
+
"learning_rate": 3.454521065542273e-05,
|
| 555 |
+
"loss": 0.4069,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.10088837146759033,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 5246.9,
|
| 560 |
+
"valid_targets_min": 1264
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.6272,
|
| 564 |
+
"grad_norm": 0.2589368745471832,
|
| 565 |
+
"learning_rate": 3.423621541950597e-05,
|
| 566 |
+
"loss": 0.3851,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.10473264008760452,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 5016.7,
|
| 571 |
+
"valid_targets_min": 1757
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.6592,
|
| 575 |
+
"grad_norm": 0.28003847497063306,
|
| 576 |
+
"learning_rate": 3.3920173139522664e-05,
|
| 577 |
+
"loss": 0.3998,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.09361633658409119,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 3909.4,
|
| 582 |
+
"valid_targets_min": 785
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.6912,
|
| 586 |
+
"grad_norm": 0.2943787611971499,
|
| 587 |
+
"learning_rate": 3.35972402590084e-05,
|
| 588 |
+
"loss": 0.3774,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.08121328055858612,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 3470.5,
|
| 593 |
+
"valid_targets_min": 780
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7231999999999998,
|
| 597 |
+
"grad_norm": 0.264050322073341,
|
| 598 |
+
"learning_rate": 3.326757663240291e-05,
|
| 599 |
+
"loss": 0.3797,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.11807336658239365,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 4398.0,
|
| 604 |
+
"valid_targets_min": 735
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7551999999999999,
|
| 608 |
+
"grad_norm": 0.2580218568904529,
|
| 609 |
+
"learning_rate": 3.293134544592073e-05,
|
| 610 |
+
"loss": 0.3618,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.096112921833992,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 5801.4,
|
| 615 |
+
"valid_targets_min": 547
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7872,
|
| 619 |
+
"grad_norm": 0.30673536056998396,
|
| 620 |
+
"learning_rate": 3.258871313677274e-05,
|
| 621 |
+
"loss": 0.3881,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.08830038458108902,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 3337.2,
|
| 626 |
+
"valid_targets_min": 1124
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8192,
|
| 630 |
+
"grad_norm": 0.2835089845775815,
|
| 631 |
+
"learning_rate": 3.2239849310778316e-05,
|
| 632 |
+
"loss": 0.4036,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.1227225810289383,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 4012.8,
|
| 637 |
+
"valid_targets_min": 861
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.8512,
|
| 641 |
+
"grad_norm": 0.2523256388128284,
|
| 642 |
+
"learning_rate": 3.188492665840909e-05,
|
| 643 |
+
"loss": 0.3794,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.08101785182952881,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 3038.2,
|
| 648 |
+
"valid_targets_min": 823
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.8832,
|
| 652 |
+
"grad_norm": 0.26924124501345237,
|
| 653 |
+
"learning_rate": 3.1524120869305726e-05,
|
| 654 |
+
"loss": 0.3806,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.10177544504404068,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 5270.2,
|
| 659 |
+
"valid_targets_min": 820
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.9152,
|
| 663 |
+
"grad_norm": 0.28130372404313303,
|
| 664 |
+
"learning_rate": 3.11576105453101e-05,
|
| 665 |
+
"loss": 0.3778,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.07558295130729675,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 4471.8,
|
| 670 |
+
"valid_targets_min": 600
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.9472,
|
| 674 |
+
"grad_norm": 0.28205364927715976,
|
| 675 |
+
"learning_rate": 3.0785577112055916e-05,
|
| 676 |
+
"loss": 0.387,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.12669092416763306,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 5147.2,
|
| 681 |
+
"valid_targets_min": 578
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9792,
|
| 685 |
+
"grad_norm": 0.272400846069077,
|
| 686 |
+
"learning_rate": 3.040820472916153e-05,
|
| 687 |
+
"loss": 0.3724,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.11067002266645432,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 3779.5,
|
| 692 |
+
"valid_targets_min": 955
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.0064,
|
| 696 |
+
"grad_norm": 0.26765963746969057,
|
| 697 |
+
"learning_rate": 3.002568019906939e-05,
|
| 698 |
+
"loss": 0.3865,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.100395068526268,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 4568.2,
|
| 703 |
+
"valid_targets_min": 792
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0384,
|
| 707 |
+
"grad_norm": 0.27706186342808137,
|
| 708 |
+
"learning_rate": 2.963819287457733e-05,
|
| 709 |
+
"loss": 0.3781,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.13206183910369873,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 4657.1,
|
| 714 |
+
"valid_targets_min": 834
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0704,
|
| 718 |
+
"grad_norm": 0.2661407560456541,
|
| 719 |
+
"learning_rate": 2.924593456510733e-05,
|
| 720 |
+
"loss": 0.3491,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.09652642160654068,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 5487.9,
|
| 725 |
+
"valid_targets_min": 1218
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.1024,
|
| 729 |
+
"grad_norm": 0.31701150346198914,
|
| 730 |
+
"learning_rate": 2.8849099441758306e-05,
|
| 731 |
+
"loss": 0.37,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.08779852092266083,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 3176.9,
|
| 736 |
+
"valid_targets_min": 762
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.1344,
|
| 740 |
+
"grad_norm": 0.27075908827848955,
|
| 741 |
+
"learning_rate": 2.844788394118979e-05,
|
| 742 |
+
"loss": 0.369,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.09712551534175873,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 4385.8,
|
| 747 |
+
"valid_targets_min": 867
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.1664,
|
| 751 |
+
"grad_norm": 0.2492639945727957,
|
| 752 |
+
"learning_rate": 2.8042486668384164e-05,
|
| 753 |
+
"loss": 0.381,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.08740992844104767,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 4696.7,
|
| 758 |
+
"valid_targets_min": 796
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.1984,
|
| 762 |
+
"grad_norm": 0.30232927389846304,
|
| 763 |
+
"learning_rate": 2.7633108298335582e-05,
|
| 764 |
+
"loss": 0.3761,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.11960571259260178,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 4552.9,
|
| 769 |
+
"valid_targets_min": 610
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.2304,
|
| 773 |
+
"grad_norm": 0.28463853698825686,
|
| 774 |
+
"learning_rate": 2.721995147671416e-05,
|
| 775 |
+
"loss": 0.3553,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.0824894830584526,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 3428.7,
|
| 780 |
+
"valid_targets_min": 645
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.2624,
|
| 784 |
+
"grad_norm": 0.2619547638651442,
|
| 785 |
+
"learning_rate": 2.68032207195547e-05,
|
| 786 |
+
"loss": 0.3674,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.12523266673088074,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 5121.7,
|
| 791 |
+
"valid_targets_min": 827
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.2944,
|
| 795 |
+
"grad_norm": 0.2826176494689443,
|
| 796 |
+
"learning_rate": 2.6383122312019604e-05,
|
| 797 |
+
"loss": 0.3698,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.10490305721759796,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 5392.1,
|
| 802 |
+
"valid_targets_min": 1163
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3264,
|
| 806 |
+
"grad_norm": 0.26281886388054504,
|
| 807 |
+
"learning_rate": 2.595986420628597e-05,
|
| 808 |
+
"loss": 0.3698,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.08223024010658264,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 4496.4,
|
| 813 |
+
"valid_targets_min": 891
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3584,
|
| 817 |
+
"grad_norm": 0.28520427596756726,
|
| 818 |
+
"learning_rate": 2.5533655918607573e-05,
|
| 819 |
+
"loss": 0.3659,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.07992172241210938,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 3386.4,
|
| 824 |
+
"valid_targets_min": 649
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.3904,
|
| 828 |
+
"grad_norm": 0.2925895212634724,
|
| 829 |
+
"learning_rate": 2.510470842560259e-05,
|
| 830 |
+
"loss": 0.3579,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.09350573271512985,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 4317.8,
|
| 835 |
+
"valid_targets_min": 546
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4224,
|
| 839 |
+
"grad_norm": 0.25993826727846614,
|
| 840 |
+
"learning_rate": 2.467323405981841e-05,
|
| 841 |
+
"loss": 0.3574,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.10308262705802917,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 5797.9,
|
| 846 |
+
"valid_targets_min": 1315
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.4544,
|
| 850 |
+
"grad_norm": 0.26929118383315276,
|
| 851 |
+
"learning_rate": 2.423944640462533e-05,
|
| 852 |
+
"loss": 0.3619,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.11039996147155762,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 4797.1,
|
| 857 |
+
"valid_targets_min": 923
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.4864,
|
| 861 |
+
"grad_norm": 0.2905654079690829,
|
| 862 |
+
"learning_rate": 2.3803560188490968e-05,
|
| 863 |
+
"loss": 0.3756,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.0772155225276947,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 2625.3,
|
| 868 |
+
"valid_targets_min": 715
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.5183999999999997,
|
| 872 |
+
"grad_norm": 0.2586265822283464,
|
| 873 |
+
"learning_rate": 2.336579117868789e-05,
|
| 874 |
+
"loss": 0.3511,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.08459736406803131,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 5125.2,
|
| 879 |
+
"valid_targets_min": 665
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5504,
|
| 883 |
+
"grad_norm": 0.24881513909392297,
|
| 884 |
+
"learning_rate": 2.292635607448711e-05,
|
| 885 |
+
"loss": 0.3666,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.1012953445315361,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 5529.9,
|
| 890 |
+
"valid_targets_min": 575
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.5824,
|
| 894 |
+
"grad_norm": 0.2477813914251035,
|
| 895 |
+
"learning_rate": 2.248547239989008e-05,
|
| 896 |
+
"loss": 0.3786,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.10448640584945679,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 5946.4,
|
| 901 |
+
"valid_targets_min": 1290
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.6144,
|
| 905 |
+
"grad_norm": 0.2861168060044399,
|
| 906 |
+
"learning_rate": 2.204335839595255e-05,
|
| 907 |
+
"loss": 0.3707,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.09149201959371567,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 4533.5,
|
| 912 |
+
"valid_targets_min": 707
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6464,
|
| 916 |
+
"grad_norm": 0.2714708833558892,
|
| 917 |
+
"learning_rate": 2.1600232912753452e-05,
|
| 918 |
+
"loss": 0.3788,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.0974833071231842,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 4389.8,
|
| 923 |
+
"valid_targets_min": 422
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.6784,
|
| 927 |
+
"grad_norm": 0.3115246578077533,
|
| 928 |
+
"learning_rate": 2.1156315301062293e-05,
|
| 929 |
+
"loss": 0.3806,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.10238995403051376,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 3236.0,
|
| 934 |
+
"valid_targets_min": 799
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.7104,
|
| 938 |
+
"grad_norm": 0.30112620981655347,
|
| 939 |
+
"learning_rate": 2.0711825303758712e-05,
|
| 940 |
+
"loss": 0.3785,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.09492282569408417,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 3743.2,
|
| 945 |
+
"valid_targets_min": 766
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7424,
|
| 949 |
+
"grad_norm": 0.2812525502947618,
|
| 950 |
+
"learning_rate": 2.0266982947057962e-05,
|
| 951 |
+
"loss": 0.3818,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.1166185587644577,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 4319.6,
|
| 956 |
+
"valid_targets_min": 1012
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.7744,
|
| 960 |
+
"grad_norm": 0.24849187677677648,
|
| 961 |
+
"learning_rate": 1.9822008431596083e-05,
|
| 962 |
+
"loss": 0.3497,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.08945291489362717,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 5814.7,
|
| 967 |
+
"valid_targets_min": 784
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.8064,
|
| 971 |
+
"grad_norm": 0.2631331467696701,
|
| 972 |
+
"learning_rate": 1.937712202342881e-05,
|
| 973 |
+
"loss": 0.3797,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.10857097804546356,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 4672.4,
|
| 978 |
+
"valid_targets_min": 1115
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8384,
|
| 982 |
+
"grad_norm": 0.26773590398116837,
|
| 983 |
+
"learning_rate": 1.8932543944998037e-05,
|
| 984 |
+
"loss": 0.3578,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.06477227061986923,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 3280.0,
|
| 989 |
+
"valid_targets_min": 463
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.8704,
|
| 993 |
+
"grad_norm": 0.2593193114272387,
|
| 994 |
+
"learning_rate": 1.8488494266119877e-05,
|
| 995 |
+
"loss": 0.3645,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.08953142166137695,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 5166.1,
|
| 1000 |
+
"valid_targets_min": 610
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.9024,
|
| 1004 |
+
"grad_norm": 0.2617695948022818,
|
| 1005 |
+
"learning_rate": 1.804519279504834e-05,
|
| 1006 |
+
"loss": 0.368,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.08089570701122284,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 3814.2,
|
| 1011 |
+
"valid_targets_min": 706
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9344,
|
| 1015 |
+
"grad_norm": 0.2357322241148187,
|
| 1016 |
+
"learning_rate": 1.7602858969668365e-05,
|
| 1017 |
+
"loss": 0.346,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.08805036544799805,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 5983.0,
|
| 1022 |
+
"valid_targets_min": 733
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.9664,
|
| 1026 |
+
"grad_norm": 0.26747148060678083,
|
| 1027 |
+
"learning_rate": 1.716171174887231e-05,
|
| 1028 |
+
"loss": 0.3777,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.11624380946159363,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 5326.9,
|
| 1033 |
+
"valid_targets_min": 807
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.9984,
|
| 1037 |
+
"grad_norm": 0.2676459351712619,
|
| 1038 |
+
"learning_rate": 1.6721969504173484e-05,
|
| 1039 |
+
"loss": 0.3754,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.10637401044368744,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 4809.9,
|
| 1044 |
+
"valid_targets_min": 683
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0256,
|
| 1048 |
+
"grad_norm": 0.2527197970631467,
|
| 1049 |
+
"learning_rate": 1.628384991161041e-05,
|
| 1050 |
+
"loss": 0.3763,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.08800717443227768,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 4288.8,
|
| 1055 |
+
"valid_targets_min": 523
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.0576,
|
| 1059 |
+
"grad_norm": 0.26943933773942935,
|
| 1060 |
+
"learning_rate": 1.5847569843995452e-05,
|
| 1061 |
+
"loss": 0.3633,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.08747033029794693,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 3784.7,
|
| 1066 |
+
"valid_targets_min": 707
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.0896,
|
| 1070 |
+
"grad_norm": 0.2929312520277543,
|
| 1071 |
+
"learning_rate": 1.5413345263560922e-05,
|
| 1072 |
+
"loss": 0.3534,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.062183164060115814,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 1691.6,
|
| 1077 |
+
"valid_targets_min": 563
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.1216,
|
| 1081 |
+
"grad_norm": 0.2637865324511221,
|
| 1082 |
+
"learning_rate": 1.4981391115056032e-05,
|
| 1083 |
+
"loss": 0.3575,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.09510040283203125,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 3853.3,
|
| 1088 |
+
"valid_targets_min": 709
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1536,
|
| 1092 |
+
"grad_norm": 0.30734784710913193,
|
| 1093 |
+
"learning_rate": 1.455192121934748e-05,
|
| 1094 |
+
"loss": 0.363,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.09190519154071808,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 3746.6,
|
| 1099 |
+
"valid_targets_min": 833
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.1856,
|
| 1103 |
+
"grad_norm": 0.24787331839944854,
|
| 1104 |
+
"learning_rate": 1.4125148167576303e-05,
|
| 1105 |
+
"loss": 0.347,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.08131758123636246,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 4313.7,
|
| 1110 |
+
"valid_targets_min": 990
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.2176,
|
| 1114 |
+
"grad_norm": 0.24383293934449407,
|
| 1115 |
+
"learning_rate": 1.3701283215923563e-05,
|
| 1116 |
+
"loss": 0.3593,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.06853271275758743,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 4191.8,
|
| 1121 |
+
"valid_targets_min": 659
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2496,
|
| 1125 |
+
"grad_norm": 0.2965221142379322,
|
| 1126 |
+
"learning_rate": 1.328053618103677e-05,
|
| 1127 |
+
"loss": 0.3468,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.07949408888816833,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 3390.6,
|
| 1132 |
+
"valid_targets_min": 762
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.2816,
|
| 1136 |
+
"grad_norm": 0.2492892033414412,
|
| 1137 |
+
"learning_rate": 1.2863115336168916e-05,
|
| 1138 |
+
"loss": 0.3505,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.07512718439102173,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 4176.8,
|
| 1143 |
+
"valid_targets_min": 830
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3136,
|
| 1147 |
+
"grad_norm": 0.267290487021705,
|
| 1148 |
+
"learning_rate": 1.2449227308081509e-05,
|
| 1149 |
+
"loss": 0.3525,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.06713395565748215,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 3405.8,
|
| 1154 |
+
"valid_targets_min": 732
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.3456,
|
| 1158 |
+
"grad_norm": 0.2551948302077571,
|
| 1159 |
+
"learning_rate": 1.2039076974762587e-05,
|
| 1160 |
+
"loss": 0.3476,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.07601924240589142,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 3619.9,
|
| 1165 |
+
"valid_targets_min": 905
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.3776,
|
| 1169 |
+
"grad_norm": 0.25252309788307625,
|
| 1170 |
+
"learning_rate": 1.163286736401044e-05,
|
| 1171 |
+
"loss": 0.3549,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.07913119345903397,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 3200.8,
|
| 1176 |
+
"valid_targets_min": 661
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4096,
|
| 1180 |
+
"grad_norm": 0.3155156588956853,
|
| 1181 |
+
"learning_rate": 1.123079955293322e-05,
|
| 1182 |
+
"loss": 0.3656,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.09287755191326141,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 3214.9,
|
| 1187 |
+
"valid_targets_min": 747
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.4416,
|
| 1191 |
+
"grad_norm": 0.2766133077094186,
|
| 1192 |
+
"learning_rate": 1.0833072568414037e-05,
|
| 1193 |
+
"loss": 0.3626,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.09605400264263153,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 4218.4,
|
| 1198 |
+
"valid_targets_min": 1095
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.4736000000000002,
|
| 1202 |
+
"grad_norm": 0.27258472577367454,
|
| 1203 |
+
"learning_rate": 1.0439883288591057e-05,
|
| 1204 |
+
"loss": 0.3696,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.08149433881044388,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 4192.5,
|
| 1209 |
+
"valid_targets_min": 747
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.5056000000000003,
|
| 1213 |
+
"grad_norm": 0.25426394436556843,
|
| 1214 |
+
"learning_rate": 1.0051426345401202e-05,
|
| 1215 |
+
"loss": 0.3589,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.08685190975666046,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 4254.9,
|
| 1220 |
+
"valid_targets_min": 757
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5376,
|
| 1224 |
+
"grad_norm": 0.27153496157844426,
|
| 1225 |
+
"learning_rate": 9.667894028235704e-06,
|
| 1226 |
+
"loss": 0.3657,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.10605305433273315,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 4210.9,
|
| 1231 |
+
"valid_targets_min": 1488
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.5696,
|
| 1235 |
+
"grad_norm": 0.3537526334104841,
|
| 1236 |
+
"learning_rate": 9.289476188755315e-06,
|
| 1237 |
+
"loss": 0.3611,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.09752020239830017,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 2972.8,
|
| 1242 |
+
"valid_targets_min": 794
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6016,
|
| 1246 |
+
"grad_norm": 0.2651341936936712,
|
| 1247 |
+
"learning_rate": 8.916360146912122e-06,
|
| 1248 |
+
"loss": 0.3505,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.07173003256320953,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 3701.8,
|
| 1253 |
+
"valid_targets_min": 734
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.6336,
|
| 1257 |
+
"grad_norm": 0.25545676170795567,
|
| 1258 |
+
"learning_rate": 8.548730598224646e-06,
|
| 1259 |
+
"loss": 0.3618,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.07571163773536682,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 4587.5,
|
| 1264 |
+
"valid_targets_min": 897
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.6656,
|
| 1268 |
+
"grad_norm": 0.25947716758369005,
|
| 1269 |
+
"learning_rate": 8.186769522352053e-06,
|
| 1270 |
+
"loss": 0.3568,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.10017608106136322,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 4887.9,
|
| 1275 |
+
"valid_targets_min": 920
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.6976,
|
| 1279 |
+
"grad_norm": 0.24666029210741366,
|
| 1280 |
+
"learning_rate": 7.830656093012714e-06,
|
| 1281 |
+
"loss": 0.3604,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.07228102535009384,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 3624.2,
|
| 1286 |
+
"valid_targets_min": 682
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.7296,
|
| 1290 |
+
"grad_norm": 0.30422926479905615,
|
| 1291 |
+
"learning_rate": 7.480566589291696e-06,
|
| 1292 |
+
"loss": 0.3529,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.0982704907655716,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 4152.2,
|
| 1297 |
+
"valid_targets_min": 872
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.7616,
|
| 1301 |
+
"grad_norm": 0.2848406589807159,
|
| 1302 |
+
"learning_rate": 7.1366743083812285e-06,
|
| 1303 |
+
"loss": 0.3565,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.06402648240327835,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 2856.8,
|
| 1308 |
+
"valid_targets_min": 1025
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.7936,
|
| 1312 |
+
"grad_norm": 0.3175945368962988,
|
| 1313 |
+
"learning_rate": 6.799149479797101e-06,
|
| 1314 |
+
"loss": 0.3625,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.07294386625289917,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 2645.3,
|
| 1319 |
+
"valid_targets_min": 1036
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.8256,
|
| 1323 |
+
"grad_norm": 0.2662096619930337,
|
| 1324 |
+
"learning_rate": 6.4681591811137e-06,
|
| 1325 |
+
"loss": 0.3669,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.07283228635787964,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 3706.9,
|
| 1330 |
+
"valid_targets_min": 1047
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.8576,
|
| 1334 |
+
"grad_norm": 0.2594850403272501,
|
| 1335 |
+
"learning_rate": 6.143867255259197e-06,
|
| 1336 |
+
"loss": 0.3504,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.10935956239700317,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 5242.9,
|
| 1341 |
+
"valid_targets_min": 665
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.8895999999999997,
|
| 1345 |
+
"grad_norm": 0.24713702492526102,
|
| 1346 |
+
"learning_rate": 5.8264342294119504e-06,
|
| 1347 |
+
"loss": 0.3458,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.06710461527109146,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 4347.1,
|
| 1352 |
+
"valid_targets_min": 1322
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.9215999999999998,
|
| 1356 |
+
"grad_norm": 0.2897155890131026,
|
| 1357 |
+
"learning_rate": 5.516017235538258e-06,
|
| 1358 |
+
"loss": 0.3625,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.08740752935409546,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 3607.1,
|
| 1363 |
+
"valid_targets_min": 600
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.9536,
|
| 1367 |
+
"grad_norm": 0.24379821604551838,
|
| 1368 |
+
"learning_rate": 5.212769932610695e-06,
|
| 1369 |
+
"loss": 0.3501,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.0911904126405716,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 4610.9,
|
| 1374 |
+
"valid_targets_min": 859
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.9856,
|
| 1378 |
+
"grad_norm": 0.3154382094028816,
|
| 1379 |
+
"learning_rate": 4.916842430545681e-06,
|
| 1380 |
+
"loss": 0.3646,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.105601966381073,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 4914.9,
|
| 1385 |
+
"valid_targets_min": 537
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.0128,
|
| 1389 |
+
"grad_norm": 0.2728978101785218,
|
| 1390 |
+
"learning_rate": 4.628381215897837e-06,
|
| 1391 |
+
"loss": 0.3477,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.06833998113870621,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 4309.9,
|
| 1396 |
+
"valid_targets_min": 612
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.0448,
|
| 1400 |
+
"grad_norm": 0.2667331818881737,
|
| 1401 |
+
"learning_rate": 4.347529079347914e-06,
|
| 1402 |
+
"loss": 0.3528,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.08249158412218094,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 3725.8,
|
| 1407 |
+
"valid_targets_min": 573
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.0768,
|
| 1411 |
+
"grad_norm": 0.2653027077039279,
|
| 1412 |
+
"learning_rate": 4.074425045020247e-06,
|
| 1413 |
+
"loss": 0.3516,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.11303485184907913,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 5256.0,
|
| 1418 |
+
"valid_targets_min": 1442
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.1088,
|
| 1422 |
+
"grad_norm": 0.26248507735732046,
|
| 1423 |
+
"learning_rate": 3.8092043016646487e-06,
|
| 1424 |
+
"loss": 0.338,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.06639324128627777,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 3318.1,
|
| 1429 |
+
"valid_targets_min": 679
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.1408,
|
| 1433 |
+
"grad_norm": 0.27929839995512357,
|
| 1434 |
+
"learning_rate": 3.551998135736867e-06,
|
| 1435 |
+
"loss": 0.3415,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.05943456292152405,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 2786.4,
|
| 1440 |
+
"valid_targets_min": 777
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.1728,
|
| 1444 |
+
"grad_norm": 0.2586565562817669,
|
| 1445 |
+
"learning_rate": 3.3029338664107267e-06,
|
| 1446 |
+
"loss": 0.3615,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.08127006888389587,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 3574.6,
|
| 1451 |
+
"valid_targets_min": 840
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.2048,
|
| 1455 |
+
"grad_norm": 0.26935723611482854,
|
| 1456 |
+
"learning_rate": 3.0621347825540625e-06,
|
| 1457 |
+
"loss": 0.3525,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.07011975347995758,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 2604.6,
|
| 1462 |
+
"valid_targets_min": 726
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.2368,
|
| 1466 |
+
"grad_norm": 0.2637733286008412,
|
| 1467 |
+
"learning_rate": 2.8297200816997183e-06,
|
| 1468 |
+
"loss": 0.3499,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.09666310250759125,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 5042.1,
|
| 1473 |
+
"valid_targets_min": 544
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.2688,
|
| 1477 |
+
"grad_norm": 0.32987725919571403,
|
| 1478 |
+
"learning_rate": 2.605804811041803e-06,
|
| 1479 |
+
"loss": 0.3476,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.08155728876590729,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 3523.2,
|
| 1484 |
+
"valid_targets_min": 953
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.3008,
|
| 1488 |
+
"grad_norm": 0.2715487393904151,
|
| 1489 |
+
"learning_rate": 2.390499810486351e-06,
|
| 1490 |
+
"loss": 0.3637,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.09598340094089508,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 4457.3,
|
| 1495 |
+
"valid_targets_min": 692
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.3328,
|
| 1499 |
+
"grad_norm": 0.2812120700507707,
|
| 1500 |
+
"learning_rate": 2.183911657784685e-06,
|
| 1501 |
+
"loss": 0.3644,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.11278092116117477,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 4421.0,
|
| 1506 |
+
"valid_targets_min": 730
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.3648,
|
| 1510 |
+
"grad_norm": 0.2579264284472369,
|
| 1511 |
+
"learning_rate": 1.986142615776532e-06,
|
| 1512 |
+
"loss": 0.35,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.07923750579357147,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 3840.6,
|
| 1517 |
+
"valid_targets_min": 759
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.3968,
|
| 1521 |
+
"grad_norm": 0.2534367716071427,
|
| 1522 |
+
"learning_rate": 1.7972905817690644e-06,
|
| 1523 |
+
"loss": 0.3572,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.06656123697757721,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 2848.0,
|
| 1528 |
+
"valid_targets_min": 892
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.4288,
|
| 1532 |
+
"grad_norm": 0.2614144350234106,
|
| 1533 |
+
"learning_rate": 1.617449039076955e-06,
|
| 1534 |
+
"loss": 0.3663,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.08406713604927063,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 4129.6,
|
| 1539 |
+
"valid_targets_min": 787
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.4608,
|
| 1543 |
+
"grad_norm": 0.26251126298421695,
|
| 1544 |
+
"learning_rate": 1.4467070107473413e-06,
|
| 1545 |
+
"loss": 0.3609,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.07219307124614716,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 3461.7,
|
| 1550 |
+
"valid_targets_min": 1063
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.4928,
|
| 1554 |
+
"grad_norm": 0.2818927782701796,
|
| 1555 |
+
"learning_rate": 1.2851490154926816e-06,
|
| 1556 |
+
"loss": 0.3505,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.14388665556907654,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 5650.9,
|
| 1561 |
+
"valid_targets_min": 662
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.5248,
|
| 1565 |
+
"grad_norm": 0.25013055924766153,
|
| 1566 |
+
"learning_rate": 1.1328550258533211e-06,
|
| 1567 |
+
"loss": 0.3463,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.078327476978302,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 3948.1,
|
| 1572 |
+
"valid_targets_min": 1196
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.5568,
|
| 1576 |
+
"grad_norm": 0.23856915999519057,
|
| 1577 |
+
"learning_rate": 9.899004286103953e-07,
|
| 1578 |
+
"loss": 0.356,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.08584430068731308,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 4553.8,
|
| 1583 |
+
"valid_targets_min": 659
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.5888,
|
| 1587 |
+
"grad_norm": 0.29180461161366616,
|
| 1588 |
+
"learning_rate": 8.5635598746876e-07,
|
| 1589 |
+
"loss": 0.3718,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.10328815877437592,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 4305.4,
|
| 1594 |
+
"valid_targets_min": 560
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.6208,
|
| 1598 |
+
"grad_norm": 0.25445248125988607,
|
| 1599 |
+
"learning_rate": 7.32287808028389e-07,
|
| 1600 |
+
"loss": 0.3412,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.08515608310699463,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 3890.6,
|
| 1605 |
+
"valid_targets_min": 784
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.6528,
|
| 1609 |
+
"grad_norm": 0.2689653858867643,
|
| 1610 |
+
"learning_rate": 6.177573050615327e-07,
|
| 1611 |
+
"loss": 0.3677,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.10580168664455414,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 3736.6,
|
| 1616 |
+
"valid_targets_min": 1083
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.6848,
|
| 1620 |
+
"grad_norm": 0.28761622388224484,
|
| 1621 |
+
"learning_rate": 5.128211721119213e-07,
|
| 1622 |
+
"loss": 0.3566,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.12114425003528595,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 4689.3,
|
| 1627 |
+
"valid_targets_min": 940
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.7168,
|
| 1631 |
+
"grad_norm": 0.24365172717782563,
|
| 1632 |
+
"learning_rate": 4.175313534309755e-07,
|
| 1633 |
+
"loss": 0.3328,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.0749347135424614,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 4185.6,
|
| 1638 |
+
"valid_targets_min": 653
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.7488,
|
| 1642 |
+
"grad_norm": 0.2283492870630903,
|
| 1643 |
+
"learning_rate": 3.319350182649861e-07,
|
| 1644 |
+
"loss": 0.3434,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.07939842343330383,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 5340.9,
|
| 1649 |
+
"valid_targets_min": 1265
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.7808,
|
| 1653 |
+
"grad_norm": 0.29806755489113096,
|
| 1654 |
+
"learning_rate": 2.560745375059392e-07,
|
| 1655 |
+
"loss": 0.3582,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.11634281277656555,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 4905.9,
|
| 1660 |
+
"valid_targets_min": 695
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.8128,
|
| 1664 |
+
"grad_norm": 0.2671590330780972,
|
| 1665 |
+
"learning_rate": 1.8998746271758016e-07,
|
| 1666 |
+
"loss": 0.3383,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.05478291213512421,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 2623.4,
|
| 1671 |
+
"valid_targets_min": 595
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.8448,
|
| 1675 |
+
"grad_norm": 0.2706109149704598,
|
| 1676 |
+
"learning_rate": 1.337065075470778e-07,
|
| 1677 |
+
"loss": 0.3525,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.07864299416542053,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 3891.2,
|
| 1682 |
+
"valid_targets_min": 668
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.8768,
|
| 1686 |
+
"grad_norm": 0.2759378816487873,
|
| 1687 |
+
"learning_rate": 8.725953153150279e-08,
|
| 1688 |
+
"loss": 0.3635,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.08948777616024017,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 3952.4,
|
| 1693 |
+
"valid_targets_min": 644
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.9088,
|
| 1697 |
+
"grad_norm": 0.24354389118800948,
|
| 1698 |
+
"learning_rate": 5.066952630711886e-08,
|
| 1699 |
+
"loss": 0.3437,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.07612688839435577,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 4240.5,
|
| 1704 |
+
"valid_targets_min": 1037
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.9408,
|
| 1708 |
+
"grad_norm": 0.2613977673526795,
|
| 1709 |
+
"learning_rate": 2.3954604228342283e-08,
|
| 1710 |
+
"loss": 0.3511,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.0692347064614296,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 3165.6,
|
| 1715 |
+
"valid_targets_min": 703
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.9728,
|
| 1719 |
+
"grad_norm": 0.25143698509914547,
|
| 1720 |
+
"learning_rate": 7.12798940197601e-09,
|
| 1721 |
+
"loss": 0.3495,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.08967740833759308,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 4549.7,
|
| 1726 |
+
"valid_targets_min": 551
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"grad_norm": 0.5494204534303122,
|
| 1731 |
+
"learning_rate": 1.9801114115480802e-10,
|
| 1732 |
+
"loss": 0.3527,
|
| 1733 |
+
"loss_nan_ranks": 0,
|
| 1734 |
+
"loss_rank_avg": 0.36648064851760864,
|
| 1735 |
+
"step": 785,
|
| 1736 |
+
"valid_targets_mean": 3627.8,
|
| 1737 |
+
"valid_targets_min": 954
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 5.0,
|
| 1741 |
+
"loss_nan_ranks": 0,
|
| 1742 |
+
"loss_rank_avg": 0.36648064851760864,
|
| 1743 |
+
"step": 785,
|
| 1744 |
+
"total_flos": 1.0910552897523548e+18,
|
| 1745 |
+
"train_loss": 0.3873556343613157,
|
| 1746 |
+
"train_runtime": 17001.6953,
|
| 1747 |
+
"train_samples_per_second": 2.941,
|
| 1748 |
+
"train_steps_per_second": 0.046,
|
| 1749 |
+
"valid_targets_mean": 3627.8,
|
| 1750 |
+
"valid_targets_min": 954
|
| 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": 1.0910552897523548e+18,
|
| 1771 |
+
"train_batch_size": 1,
|
| 1772 |
+
"trial_name": null,
|
| 1773 |
+
"trial_params": null
|
| 1774 |
+
}
|
training_loss.png
ADDED
|