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_seed_45
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# code-contests-sandboxes-traces-terminus-2_seed_45
|
| 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_seed_45
|
|
|
|
| 16 |
|
| 17 |
# code-contests-sandboxes-traces-terminus-2_seed_45
|
| 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": 3.938212952590612,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 152.45570532704255,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.34793925285339355,
|
| 7 |
+
"mfu_percent": 0.2783189365788419,
|
| 8 |
+
"mfu_percent_theoretical": 10.77425479343057,
|
| 9 |
+
"total_flos": 1.0929687562016522e+18,
|
| 10 |
+
"train_loss": 0.3876832940775877,
|
| 11 |
+
"train_runtime": 17345.5697,
|
| 12 |
+
"train_samples_per_second": 2.883,
|
| 13 |
+
"train_steps_per_second": 0.045,
|
| 14 |
+
"valid_targets_mean": 3301.4,
|
| 15 |
+
"valid_targets_min": 867
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 3.938212952590612,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 152.45570532704255,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.34793925285339355,
|
| 7 |
+
"mfu_percent": 0.2783189365788419,
|
| 8 |
+
"mfu_percent_theoretical": 10.77425479343057,
|
| 9 |
+
"total_flos": 1.0929687562016522e+18,
|
| 10 |
+
"train_loss": 0.3876832940775877,
|
| 11 |
+
"train_runtime": 17345.5697,
|
| 12 |
+
"train_samples_per_second": 2.883,
|
| 13 |
+
"train_steps_per_second": 0.045,
|
| 14 |
+
"valid_targets_mean": 3301.4,
|
| 15 |
+
"valid_targets_min": 867
|
| 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": 5.541497684638077,
|
| 15 |
+
"learning_rate": 2.0253164556962026e-06,
|
| 16 |
+
"loss": 0.7201,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.20204728841781616,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 3694.1,
|
| 21 |
+
"valid_targets_min": 918
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.064,
|
| 25 |
+
"grad_norm": 3.5800693611841905,
|
| 26 |
+
"learning_rate": 4.556962025316456e-06,
|
| 27 |
+
"loss": 0.6795,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.1315670758485794,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 4516.5,
|
| 32 |
+
"valid_targets_min": 765
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.096,
|
| 36 |
+
"grad_norm": 1.9884248548900978,
|
| 37 |
+
"learning_rate": 7.08860759493671e-06,
|
| 38 |
+
"loss": 0.6312,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.13017678260803223,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 4166.5,
|
| 43 |
+
"valid_targets_min": 845
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.128,
|
| 47 |
+
"grad_norm": 0.9233035770274349,
|
| 48 |
+
"learning_rate": 9.620253164556963e-06,
|
| 49 |
+
"loss": 0.6036,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.18069788813591003,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 5295.3,
|
| 54 |
+
"valid_targets_min": 864
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.16,
|
| 58 |
+
"grad_norm": 0.725540551324882,
|
| 59 |
+
"learning_rate": 1.2151898734177216e-05,
|
| 60 |
+
"loss": 0.5716,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.1643570065498352,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 5433.7,
|
| 65 |
+
"valid_targets_min": 1289
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.192,
|
| 69 |
+
"grad_norm": 0.5509203398092252,
|
| 70 |
+
"learning_rate": 1.468354430379747e-05,
|
| 71 |
+
"loss": 0.5351,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.11187565326690674,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 2928.9,
|
| 76 |
+
"valid_targets_min": 793
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.224,
|
| 80 |
+
"grad_norm": 0.4718786929689295,
|
| 81 |
+
"learning_rate": 1.7215189873417723e-05,
|
| 82 |
+
"loss": 0.5297,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.15530270338058472,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 4836.8,
|
| 87 |
+
"valid_targets_min": 881
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.256,
|
| 91 |
+
"grad_norm": 0.4068295200643166,
|
| 92 |
+
"learning_rate": 1.974683544303798e-05,
|
| 93 |
+
"loss": 0.4873,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.12239599227905273,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 3672.7,
|
| 98 |
+
"valid_targets_min": 1172
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.288,
|
| 102 |
+
"grad_norm": 0.34726917547603475,
|
| 103 |
+
"learning_rate": 2.2278481012658228e-05,
|
| 104 |
+
"loss": 0.4779,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.12165165692567825,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 4045.2,
|
| 109 |
+
"valid_targets_min": 834
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32,
|
| 113 |
+
"grad_norm": 0.29930653523131256,
|
| 114 |
+
"learning_rate": 2.481012658227848e-05,
|
| 115 |
+
"loss": 0.4638,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.10612066090106964,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 2865.4,
|
| 120 |
+
"valid_targets_min": 630
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.352,
|
| 124 |
+
"grad_norm": 0.3534925450119451,
|
| 125 |
+
"learning_rate": 2.7341772151898737e-05,
|
| 126 |
+
"loss": 0.4716,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.12155529856681824,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 3224.7,
|
| 131 |
+
"valid_targets_min": 530
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.384,
|
| 135 |
+
"grad_norm": 0.2698100119387162,
|
| 136 |
+
"learning_rate": 2.987341772151899e-05,
|
| 137 |
+
"loss": 0.4361,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.07520334422588348,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 3178.1,
|
| 142 |
+
"valid_targets_min": 672
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.416,
|
| 146 |
+
"grad_norm": 0.31765210809501726,
|
| 147 |
+
"learning_rate": 3.240506329113924e-05,
|
| 148 |
+
"loss": 0.4641,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.12320488691329956,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 4320.3,
|
| 153 |
+
"valid_targets_min": 814
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.448,
|
| 157 |
+
"grad_norm": 0.2981457203142397,
|
| 158 |
+
"learning_rate": 3.49367088607595e-05,
|
| 159 |
+
"loss": 0.457,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.09969944506883621,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 2920.1,
|
| 164 |
+
"valid_targets_min": 701
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48,
|
| 168 |
+
"grad_norm": 0.2521133577518038,
|
| 169 |
+
"learning_rate": 3.746835443037975e-05,
|
| 170 |
+
"loss": 0.4469,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.10771138221025467,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 4220.4,
|
| 175 |
+
"valid_targets_min": 902
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.512,
|
| 179 |
+
"grad_norm": 0.34779990596233723,
|
| 180 |
+
"learning_rate": 4e-05,
|
| 181 |
+
"loss": 0.4457,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.141748309135437,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 3178.2,
|
| 186 |
+
"valid_targets_min": 619
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.544,
|
| 190 |
+
"grad_norm": 0.2666291404893624,
|
| 191 |
+
"learning_rate": 3.999504991751045e-05,
|
| 192 |
+
"loss": 0.431,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.112694151699543,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 5044.8,
|
| 197 |
+
"valid_targets_min": 1099
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.576,
|
| 201 |
+
"grad_norm": 0.2824399829410414,
|
| 202 |
+
"learning_rate": 3.9980202120373464e-05,
|
| 203 |
+
"loss": 0.4267,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.10418764501810074,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 3889.9,
|
| 208 |
+
"valid_targets_min": 699
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.608,
|
| 212 |
+
"grad_norm": 0.31908682280434025,
|
| 213 |
+
"learning_rate": 3.995546395837111e-05,
|
| 214 |
+
"loss": 0.4239,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.09118764102458954,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 2867.9,
|
| 219 |
+
"valid_targets_min": 850
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.64,
|
| 223 |
+
"grad_norm": 0.2549712848477804,
|
| 224 |
+
"learning_rate": 3.992084767709763e-05,
|
| 225 |
+
"loss": 0.4125,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.09206928312778473,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 4052.6,
|
| 230 |
+
"valid_targets_min": 491
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.672,
|
| 234 |
+
"grad_norm": 0.45582965229062333,
|
| 235 |
+
"learning_rate": 3.987637041189781e-05,
|
| 236 |
+
"loss": 0.4312,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.09463967382907867,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 2923.5,
|
| 241 |
+
"valid_targets_min": 830
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.704,
|
| 245 |
+
"grad_norm": 0.2707218225881968,
|
| 246 |
+
"learning_rate": 3.982205417938482e-05,
|
| 247 |
+
"loss": 0.4085,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.09323331713676453,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 4330.3,
|
| 252 |
+
"valid_targets_min": 723
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.736,
|
| 256 |
+
"grad_norm": 0.3153755815756895,
|
| 257 |
+
"learning_rate": 3.975792586654179e-05,
|
| 258 |
+
"loss": 0.4184,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.1101507917046547,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 4692.8,
|
| 263 |
+
"valid_targets_min": 711
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.768,
|
| 267 |
+
"grad_norm": 0.3295458677006816,
|
| 268 |
+
"learning_rate": 3.968401721741259e-05,
|
| 269 |
+
"loss": 0.408,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.08722187578678131,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 2657.8,
|
| 274 |
+
"valid_targets_min": 813
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8,
|
| 278 |
+
"grad_norm": 0.3038603723433472,
|
| 279 |
+
"learning_rate": 3.960036481738819e-05,
|
| 280 |
+
"loss": 0.4321,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.11620064079761505,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 3377.4,
|
| 285 |
+
"valid_targets_min": 634
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.832,
|
| 289 |
+
"grad_norm": 0.28009890305331936,
|
| 290 |
+
"learning_rate": 3.950701007509667e-05,
|
| 291 |
+
"loss": 0.4124,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.09834864735603333,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 3717.2,
|
| 296 |
+
"valid_targets_min": 695
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.864,
|
| 300 |
+
"grad_norm": 0.2566202226762974,
|
| 301 |
+
"learning_rate": 3.940399920190552e-05,
|
| 302 |
+
"loss": 0.3932,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.07297708839178085,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 2894.3,
|
| 307 |
+
"valid_targets_min": 463
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.896,
|
| 311 |
+
"grad_norm": 0.31636429084131246,
|
| 312 |
+
"learning_rate": 3.92913831890467e-05,
|
| 313 |
+
"loss": 0.4012,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.08311374485492706,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 4278.8,
|
| 318 |
+
"valid_targets_min": 812
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.928,
|
| 322 |
+
"grad_norm": 0.2576702240508463,
|
| 323 |
+
"learning_rate": 3.916921778237556e-05,
|
| 324 |
+
"loss": 0.4093,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.10160231590270996,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 4623.7,
|
| 329 |
+
"valid_targets_min": 662
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.96,
|
| 333 |
+
"grad_norm": 0.2877741398086508,
|
| 334 |
+
"learning_rate": 3.903756345477612e-05,
|
| 335 |
+
"loss": 0.4071,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.1042058914899826,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 3411.9,
|
| 340 |
+
"valid_targets_min": 596
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.992,
|
| 344 |
+
"grad_norm": 0.2938652286141047,
|
| 345 |
+
"learning_rate": 3.889648537622657e-05,
|
| 346 |
+
"loss": 0.4119,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.10854722559452057,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 3161.6,
|
| 351 |
+
"valid_targets_min": 555
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0192,
|
| 355 |
+
"grad_norm": 0.33200094179793993,
|
| 356 |
+
"learning_rate": 3.874605338153952e-05,
|
| 357 |
+
"loss": 0.3916,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.08427420258522034,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 2848.1,
|
| 362 |
+
"valid_targets_min": 719
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0512,
|
| 366 |
+
"grad_norm": 0.26215404770192946,
|
| 367 |
+
"learning_rate": 3.8586341935793265e-05,
|
| 368 |
+
"loss": 0.3837,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.06924141198396683,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 2998.8,
|
| 373 |
+
"valid_targets_min": 1000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0832,
|
| 377 |
+
"grad_norm": 0.3016475809185609,
|
| 378 |
+
"learning_rate": 3.841743009747089e-05,
|
| 379 |
+
"loss": 0.3976,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.07719582319259644,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 2974.9,
|
| 384 |
+
"valid_targets_min": 907
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1152,
|
| 388 |
+
"grad_norm": 0.3047060279415802,
|
| 389 |
+
"learning_rate": 3.8239401479325714e-05,
|
| 390 |
+
"loss": 0.3772,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.0757255107164383,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 2829.6,
|
| 395 |
+
"valid_targets_min": 764
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1472,
|
| 399 |
+
"grad_norm": 0.3273636057159624,
|
| 400 |
+
"learning_rate": 3.8052344206992276e-05,
|
| 401 |
+
"loss": 0.3891,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.1268865019083023,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 4603.5,
|
| 406 |
+
"valid_targets_min": 723
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1792,
|
| 410 |
+
"grad_norm": 0.3125871525191448,
|
| 411 |
+
"learning_rate": 3.7856350875363396e-05,
|
| 412 |
+
"loss": 0.4073,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.12131393700838089,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 4125.5,
|
| 417 |
+
"valid_targets_min": 639
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2112,
|
| 421 |
+
"grad_norm": 0.32796667514693834,
|
| 422 |
+
"learning_rate": 3.765151850275497e-05,
|
| 423 |
+
"loss": 0.3852,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.10205736756324768,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 3710.7,
|
| 428 |
+
"valid_targets_min": 878
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.2432,
|
| 432 |
+
"grad_norm": 0.24475735520158587,
|
| 433 |
+
"learning_rate": 3.7437948482881104e-05,
|
| 434 |
+
"loss": 0.3728,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.10854168981313705,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 6517.8,
|
| 439 |
+
"valid_targets_min": 837
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2752,
|
| 443 |
+
"grad_norm": 0.8583151313832471,
|
| 444 |
+
"learning_rate": 3.721574653466336e-05,
|
| 445 |
+
"loss": 0.3954,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.07846558094024658,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 2350.1,
|
| 450 |
+
"valid_targets_min": 732
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3072,
|
| 454 |
+
"grad_norm": 0.30922015694302607,
|
| 455 |
+
"learning_rate": 3.698502264989903e-05,
|
| 456 |
+
"loss": 0.4072,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.08257310092449188,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 3387.9,
|
| 461 |
+
"valid_targets_min": 864
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.3392,
|
| 465 |
+
"grad_norm": 0.25189684546137714,
|
| 466 |
+
"learning_rate": 3.674589103881432e-05,
|
| 467 |
+
"loss": 0.3993,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.09581450372934341,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 3629.7,
|
| 472 |
+
"valid_targets_min": 731
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3712,
|
| 476 |
+
"grad_norm": 0.29112659892347204,
|
| 477 |
+
"learning_rate": 3.64984700735293e-05,
|
| 478 |
+
"loss": 0.3882,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.06733721494674683,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 2276.2,
|
| 483 |
+
"valid_targets_min": 667
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4032,
|
| 487 |
+
"grad_norm": 0.30010766829015695,
|
| 488 |
+
"learning_rate": 3.624288222946273e-05,
|
| 489 |
+
"loss": 0.4001,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.1396431028842926,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 4677.5,
|
| 494 |
+
"valid_targets_min": 672
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4352,
|
| 498 |
+
"grad_norm": 0.33649560235455167,
|
| 499 |
+
"learning_rate": 3.597925402470578e-05,
|
| 500 |
+
"loss": 0.3982,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.07678442448377609,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 2519.9,
|
| 505 |
+
"valid_targets_min": 898
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4672,
|
| 509 |
+
"grad_norm": 0.2987276715695913,
|
| 510 |
+
"learning_rate": 3.570771595739445e-05,
|
| 511 |
+
"loss": 0.3941,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.10258451104164124,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 3978.2,
|
| 516 |
+
"valid_targets_min": 762
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.4992,
|
| 520 |
+
"grad_norm": 0.26735894969723445,
|
| 521 |
+
"learning_rate": 3.5428402441111964e-05,
|
| 522 |
+
"loss": 0.3856,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.09502792358398438,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 5169.4,
|
| 527 |
+
"valid_targets_min": 662
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5312000000000001,
|
| 531 |
+
"grad_norm": 0.2790211030720928,
|
| 532 |
+
"learning_rate": 3.5141451738352936e-05,
|
| 533 |
+
"loss": 0.3793,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.08720822632312775,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 3078.2,
|
| 538 |
+
"valid_targets_min": 1129
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.5632000000000001,
|
| 542 |
+
"grad_norm": 0.28420101467873393,
|
| 543 |
+
"learning_rate": 3.4847005892082266e-05,
|
| 544 |
+
"loss": 0.3923,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.06919790804386139,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 2619.1,
|
| 549 |
+
"valid_targets_min": 686
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.5952,
|
| 553 |
+
"grad_norm": 0.28761286823955284,
|
| 554 |
+
"learning_rate": 3.454521065542273e-05,
|
| 555 |
+
"loss": 0.3985,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.09496282041072845,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 5359.2,
|
| 560 |
+
"valid_targets_min": 701
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.6272,
|
| 564 |
+
"grad_norm": 0.28798208612082316,
|
| 565 |
+
"learning_rate": 3.423621541950597e-05,
|
| 566 |
+
"loss": 0.3753,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.1058167815208435,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 4252.7,
|
| 571 |
+
"valid_targets_min": 869
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.6592,
|
| 575 |
+
"grad_norm": 0.2905843016289666,
|
| 576 |
+
"learning_rate": 3.3920173139522664e-05,
|
| 577 |
+
"loss": 0.3973,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.07735498249530792,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 3003.8,
|
| 582 |
+
"valid_targets_min": 528
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.6912,
|
| 586 |
+
"grad_norm": 0.2399842469644208,
|
| 587 |
+
"learning_rate": 3.35972402590084e-05,
|
| 588 |
+
"loss": 0.3856,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.07514525949954987,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 4324.5,
|
| 593 |
+
"valid_targets_min": 711
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7231999999999998,
|
| 597 |
+
"grad_norm": 0.2769708989910312,
|
| 598 |
+
"learning_rate": 3.326757663240291e-05,
|
| 599 |
+
"loss": 0.3741,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.08158373832702637,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 3444.5,
|
| 604 |
+
"valid_targets_min": 832
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7551999999999999,
|
| 608 |
+
"grad_norm": 0.2865165779666472,
|
| 609 |
+
"learning_rate": 3.293134544592073e-05,
|
| 610 |
+
"loss": 0.3742,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.10593311488628387,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 4072.2,
|
| 615 |
+
"valid_targets_min": 753
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7872,
|
| 619 |
+
"grad_norm": 0.25372481435497135,
|
| 620 |
+
"learning_rate": 3.258871313677274e-05,
|
| 621 |
+
"loss": 0.3813,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.10056015104055405,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 4768.8,
|
| 626 |
+
"valid_targets_min": 1129
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8192,
|
| 630 |
+
"grad_norm": 0.2678766717461159,
|
| 631 |
+
"learning_rate": 3.2239849310778316e-05,
|
| 632 |
+
"loss": 0.3697,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.0960194319486618,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 4018.4,
|
| 637 |
+
"valid_targets_min": 684
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.8512,
|
| 641 |
+
"grad_norm": 0.26041593540773744,
|
| 642 |
+
"learning_rate": 3.188492665840909e-05,
|
| 643 |
+
"loss": 0.3824,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.10017910599708557,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 4154.5,
|
| 648 |
+
"valid_targets_min": 636
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.8832,
|
| 652 |
+
"grad_norm": 0.26247724384317117,
|
| 653 |
+
"learning_rate": 3.1524120869305726e-05,
|
| 654 |
+
"loss": 0.3992,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.07477947324514389,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 3326.6,
|
| 659 |
+
"valid_targets_min": 815
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.9152,
|
| 663 |
+
"grad_norm": 0.2498670350702193,
|
| 664 |
+
"learning_rate": 3.11576105453101e-05,
|
| 665 |
+
"loss": 0.3685,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.09179487824440002,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 4956.8,
|
| 670 |
+
"valid_targets_min": 819
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.9472,
|
| 674 |
+
"grad_norm": 0.2369503923852665,
|
| 675 |
+
"learning_rate": 3.0785577112055916e-05,
|
| 676 |
+
"loss": 0.3797,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.09301997721195221,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 5283.4,
|
| 681 |
+
"valid_targets_min": 711
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9792,
|
| 685 |
+
"grad_norm": 0.28546224115455,
|
| 686 |
+
"learning_rate": 3.040820472916153e-05,
|
| 687 |
+
"loss": 0.379,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.1093452051281929,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 5712.8,
|
| 692 |
+
"valid_targets_min": 623
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.0064,
|
| 696 |
+
"grad_norm": 0.2957261573831895,
|
| 697 |
+
"learning_rate": 3.002568019906939e-05,
|
| 698 |
+
"loss": 0.3803,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.073640838265419,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 2807.1,
|
| 703 |
+
"valid_targets_min": 676
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0384,
|
| 707 |
+
"grad_norm": 0.2797029601720705,
|
| 708 |
+
"learning_rate": 2.963819287457733e-05,
|
| 709 |
+
"loss": 0.3644,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.0723484680056572,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 4557.2,
|
| 714 |
+
"valid_targets_min": 626
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0704,
|
| 718 |
+
"grad_norm": 0.25397901615246865,
|
| 719 |
+
"learning_rate": 2.924593456510733e-05,
|
| 720 |
+
"loss": 0.3755,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.09201750159263611,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 4751.2,
|
| 725 |
+
"valid_targets_min": 800
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.1024,
|
| 729 |
+
"grad_norm": 0.2755393810570339,
|
| 730 |
+
"learning_rate": 2.8849099441758306e-05,
|
| 731 |
+
"loss": 0.3752,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.09428532421588898,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 4889.7,
|
| 736 |
+
"valid_targets_min": 817
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.1344,
|
| 740 |
+
"grad_norm": 0.32318560777114086,
|
| 741 |
+
"learning_rate": 2.844788394118979e-05,
|
| 742 |
+
"loss": 0.3801,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.14423143863677979,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 3977.7,
|
| 747 |
+
"valid_targets_min": 613
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.1664,
|
| 751 |
+
"grad_norm": 0.24158332518684553,
|
| 752 |
+
"learning_rate": 2.8042486668384164e-05,
|
| 753 |
+
"loss": 0.3655,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.05750841647386551,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 2743.6,
|
| 758 |
+
"valid_targets_min": 767
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.1984,
|
| 762 |
+
"grad_norm": 0.2828544403010422,
|
| 763 |
+
"learning_rate": 2.7633108298335582e-05,
|
| 764 |
+
"loss": 0.3644,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.09143859148025513,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 3692.8,
|
| 769 |
+
"valid_targets_min": 679
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.2304,
|
| 773 |
+
"grad_norm": 0.30087857652855765,
|
| 774 |
+
"learning_rate": 2.721995147671416e-05,
|
| 775 |
+
"loss": 0.3843,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.10722360759973526,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 4067.9,
|
| 780 |
+
"valid_targets_min": 749
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.2624,
|
| 784 |
+
"grad_norm": 0.27604332089737327,
|
| 785 |
+
"learning_rate": 2.68032207195547e-05,
|
| 786 |
+
"loss": 0.3884,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.09173067659139633,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 4361.7,
|
| 791 |
+
"valid_targets_min": 761
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.2944,
|
| 795 |
+
"grad_norm": 0.41830966168714057,
|
| 796 |
+
"learning_rate": 2.6383122312019604e-05,
|
| 797 |
+
"loss": 0.3519,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.08942142128944397,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 4370.5,
|
| 802 |
+
"valid_targets_min": 955
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3264,
|
| 806 |
+
"grad_norm": 0.26532006086060894,
|
| 807 |
+
"learning_rate": 2.595986420628597e-05,
|
| 808 |
+
"loss": 0.3687,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.08756840229034424,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 3483.9,
|
| 813 |
+
"valid_targets_min": 709
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3584,
|
| 817 |
+
"grad_norm": 0.25399461718436944,
|
| 818 |
+
"learning_rate": 2.5533655918607573e-05,
|
| 819 |
+
"loss": 0.3553,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.06391710042953491,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 3258.9,
|
| 824 |
+
"valid_targets_min": 762
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.3904,
|
| 828 |
+
"grad_norm": 0.26459151394063957,
|
| 829 |
+
"learning_rate": 2.510470842560259e-05,
|
| 830 |
+
"loss": 0.3797,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.09428556263446808,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 3723.7,
|
| 835 |
+
"valid_targets_min": 554
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4224,
|
| 839 |
+
"grad_norm": 0.2592657995166384,
|
| 840 |
+
"learning_rate": 2.467323405981841e-05,
|
| 841 |
+
"loss": 0.367,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.10876058042049408,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 5299.8,
|
| 846 |
+
"valid_targets_min": 820
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.4544,
|
| 850 |
+
"grad_norm": 0.3133546870917826,
|
| 851 |
+
"learning_rate": 2.423944640462533e-05,
|
| 852 |
+
"loss": 0.3883,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.09236639738082886,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 3117.1,
|
| 857 |
+
"valid_targets_min": 682
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.4864,
|
| 861 |
+
"grad_norm": 0.2805498780921623,
|
| 862 |
+
"learning_rate": 2.3803560188490968e-05,
|
| 863 |
+
"loss": 0.3518,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.11789987981319427,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 5694.6,
|
| 868 |
+
"valid_targets_min": 657
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.5183999999999997,
|
| 872 |
+
"grad_norm": 0.2468211553990243,
|
| 873 |
+
"learning_rate": 2.336579117868789e-05,
|
| 874 |
+
"loss": 0.3637,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.07164323329925537,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 3688.2,
|
| 879 |
+
"valid_targets_min": 734
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5504,
|
| 883 |
+
"grad_norm": 0.25507421149574355,
|
| 884 |
+
"learning_rate": 2.292635607448711e-05,
|
| 885 |
+
"loss": 0.3798,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.08426293730735779,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 4809.1,
|
| 890 |
+
"valid_targets_min": 844
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.5824,
|
| 894 |
+
"grad_norm": 0.25872300499927087,
|
| 895 |
+
"learning_rate": 2.248547239989008e-05,
|
| 896 |
+
"loss": 0.3658,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.09084966778755188,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 3788.1,
|
| 901 |
+
"valid_targets_min": 596
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.6144,
|
| 905 |
+
"grad_norm": 0.24955127728931853,
|
| 906 |
+
"learning_rate": 2.204335839595255e-05,
|
| 907 |
+
"loss": 0.3499,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.07429323345422745,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 4410.4,
|
| 912 |
+
"valid_targets_min": 677
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6464,
|
| 916 |
+
"grad_norm": 0.28017640413377953,
|
| 917 |
+
"learning_rate": 2.1600232912753452e-05,
|
| 918 |
+
"loss": 0.369,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.0956093892455101,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 4106.1,
|
| 923 |
+
"valid_targets_min": 1154
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.6784,
|
| 927 |
+
"grad_norm": 0.2532898705270473,
|
| 928 |
+
"learning_rate": 2.1156315301062293e-05,
|
| 929 |
+
"loss": 0.3711,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.1108713299036026,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 5110.1,
|
| 934 |
+
"valid_targets_min": 949
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.7104,
|
| 938 |
+
"grad_norm": 0.2730879181166748,
|
| 939 |
+
"learning_rate": 2.0711825303758712e-05,
|
| 940 |
+
"loss": 0.3695,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.1020272970199585,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 4568.8,
|
| 945 |
+
"valid_targets_min": 758
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7424,
|
| 949 |
+
"grad_norm": 0.2832238642802859,
|
| 950 |
+
"learning_rate": 2.0266982947057962e-05,
|
| 951 |
+
"loss": 0.3702,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.08114699274301529,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 2751.1,
|
| 956 |
+
"valid_targets_min": 805
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.7744,
|
| 960 |
+
"grad_norm": 0.29396256819439887,
|
| 961 |
+
"learning_rate": 1.9822008431596083e-05,
|
| 962 |
+
"loss": 0.3769,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.08736386895179749,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 4836.0,
|
| 967 |
+
"valid_targets_min": 824
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.8064,
|
| 971 |
+
"grad_norm": 0.293725631024192,
|
| 972 |
+
"learning_rate": 1.937712202342881e-05,
|
| 973 |
+
"loss": 0.3629,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.05763329192996025,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 2313.6,
|
| 978 |
+
"valid_targets_min": 765
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8384,
|
| 982 |
+
"grad_norm": 0.2877941893706849,
|
| 983 |
+
"learning_rate": 1.8932543944998037e-05,
|
| 984 |
+
"loss": 0.3755,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.09069660305976868,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 3941.3,
|
| 989 |
+
"valid_targets_min": 739
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.8704,
|
| 993 |
+
"grad_norm": 0.29559513503682666,
|
| 994 |
+
"learning_rate": 1.8488494266119877e-05,
|
| 995 |
+
"loss": 0.3465,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.08671572804450989,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 4384.7,
|
| 1000 |
+
"valid_targets_min": 869
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.9024,
|
| 1004 |
+
"grad_norm": 0.2926583293018059,
|
| 1005 |
+
"learning_rate": 1.804519279504834e-05,
|
| 1006 |
+
"loss": 0.3628,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.07952509820461273,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 4060.8,
|
| 1011 |
+
"valid_targets_min": 716
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9344,
|
| 1015 |
+
"grad_norm": 0.27213293831765606,
|
| 1016 |
+
"learning_rate": 1.7602858969668365e-05,
|
| 1017 |
+
"loss": 0.3602,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.09436945617198944,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 3749.9,
|
| 1022 |
+
"valid_targets_min": 602
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.9664,
|
| 1026 |
+
"grad_norm": 0.26583061680884507,
|
| 1027 |
+
"learning_rate": 1.716171174887231e-05,
|
| 1028 |
+
"loss": 0.3725,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.0769558995962143,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 3624.6,
|
| 1033 |
+
"valid_targets_min": 557
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 2.9984,
|
| 1037 |
+
"grad_norm": 0.28691255129637394,
|
| 1038 |
+
"learning_rate": 1.6721969504173484e-05,
|
| 1039 |
+
"loss": 0.3695,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.0934470146894455,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 3964.6,
|
| 1044 |
+
"valid_targets_min": 946
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0256,
|
| 1048 |
+
"grad_norm": 0.27694974177269244,
|
| 1049 |
+
"learning_rate": 1.628384991161041e-05,
|
| 1050 |
+
"loss": 0.369,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.07107527554035187,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 3270.2,
|
| 1055 |
+
"valid_targets_min": 705
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.0576,
|
| 1059 |
+
"grad_norm": 0.2735034341978741,
|
| 1060 |
+
"learning_rate": 1.5847569843995452e-05,
|
| 1061 |
+
"loss": 0.3475,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.0945439338684082,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 4498.7,
|
| 1066 |
+
"valid_targets_min": 275
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.0896,
|
| 1070 |
+
"grad_norm": 0.23880890297429838,
|
| 1071 |
+
"learning_rate": 1.5413345263560922e-05,
|
| 1072 |
+
"loss": 0.3451,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.0911998599767685,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 4708.6,
|
| 1077 |
+
"valid_targets_min": 739
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.1216,
|
| 1081 |
+
"grad_norm": 0.27371534202194003,
|
| 1082 |
+
"learning_rate": 1.4981391115056032e-05,
|
| 1083 |
+
"loss": 0.3649,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.09913412481546402,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 4566.2,
|
| 1088 |
+
"valid_targets_min": 661
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1536,
|
| 1092 |
+
"grad_norm": 0.27627616833465707,
|
| 1093 |
+
"learning_rate": 1.455192121934748e-05,
|
| 1094 |
+
"loss": 0.3702,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.09827420115470886,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 4863.6,
|
| 1099 |
+
"valid_targets_min": 780
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.1856,
|
| 1103 |
+
"grad_norm": 0.30942075492779997,
|
| 1104 |
+
"learning_rate": 1.4125148167576303e-05,
|
| 1105 |
+
"loss": 0.3541,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.1027892604470253,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 3456.2,
|
| 1110 |
+
"valid_targets_min": 737
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.2176,
|
| 1114 |
+
"grad_norm": 0.3011990578205873,
|
| 1115 |
+
"learning_rate": 1.3701283215923563e-05,
|
| 1116 |
+
"loss": 0.3848,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.10471899807453156,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 3511.1,
|
| 1121 |
+
"valid_targets_min": 1126
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2496,
|
| 1125 |
+
"grad_norm": 0.25586343923489,
|
| 1126 |
+
"learning_rate": 1.328053618103677e-05,
|
| 1127 |
+
"loss": 0.3531,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.06394007802009583,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 3619.6,
|
| 1132 |
+
"valid_targets_min": 801
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.2816,
|
| 1136 |
+
"grad_norm": 0.247916108494188,
|
| 1137 |
+
"learning_rate": 1.2863115336168916e-05,
|
| 1138 |
+
"loss": 0.3519,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.06505729258060455,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 3381.6,
|
| 1143 |
+
"valid_targets_min": 870
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3136,
|
| 1147 |
+
"grad_norm": 0.28193646168227443,
|
| 1148 |
+
"learning_rate": 1.2449227308081509e-05,
|
| 1149 |
+
"loss": 0.3641,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.07794380187988281,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 3805.7,
|
| 1154 |
+
"valid_targets_min": 808
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.3456,
|
| 1158 |
+
"grad_norm": 0.25690023433577935,
|
| 1159 |
+
"learning_rate": 1.2039076974762587e-05,
|
| 1160 |
+
"loss": 0.3456,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.0881698876619339,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 4807.1,
|
| 1165 |
+
"valid_targets_min": 665
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.3776,
|
| 1169 |
+
"grad_norm": 0.2549757887509029,
|
| 1170 |
+
"learning_rate": 1.163286736401044e-05,
|
| 1171 |
+
"loss": 0.3581,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.07912416756153107,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 3860.9,
|
| 1176 |
+
"valid_targets_min": 596
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4096,
|
| 1180 |
+
"grad_norm": 0.28204253021652476,
|
| 1181 |
+
"learning_rate": 1.123079955293322e-05,
|
| 1182 |
+
"loss": 0.3515,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.09103786200284958,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 3632.4,
|
| 1187 |
+
"valid_targets_min": 833
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.4416,
|
| 1191 |
+
"grad_norm": 0.2943012232723084,
|
| 1192 |
+
"learning_rate": 1.0833072568414037e-05,
|
| 1193 |
+
"loss": 0.3502,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.10494515299797058,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 4191.4,
|
| 1198 |
+
"valid_targets_min": 760
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.4736000000000002,
|
| 1202 |
+
"grad_norm": 0.2601962964616993,
|
| 1203 |
+
"learning_rate": 1.0439883288591057e-05,
|
| 1204 |
+
"loss": 0.361,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.11226920038461685,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 5768.5,
|
| 1209 |
+
"valid_targets_min": 1032
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.5056000000000003,
|
| 1213 |
+
"grad_norm": 0.2771216636065589,
|
| 1214 |
+
"learning_rate": 1.0051426345401202e-05,
|
| 1215 |
+
"loss": 0.3528,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.0983453094959259,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 4185.4,
|
| 1220 |
+
"valid_targets_min": 1072
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5376,
|
| 1224 |
+
"grad_norm": 0.29687442352238086,
|
| 1225 |
+
"learning_rate": 9.667894028235704e-06,
|
| 1226 |
+
"loss": 0.3695,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.10656087100505829,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 3893.4,
|
| 1231 |
+
"valid_targets_min": 586
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.5696,
|
| 1235 |
+
"grad_norm": 0.3258078619224109,
|
| 1236 |
+
"learning_rate": 9.289476188755315e-06,
|
| 1237 |
+
"loss": 0.3536,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.09270378947257996,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 3400.8,
|
| 1242 |
+
"valid_targets_min": 671
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6016,
|
| 1246 |
+
"grad_norm": 0.25100309171961954,
|
| 1247 |
+
"learning_rate": 8.916360146912122e-06,
|
| 1248 |
+
"loss": 0.3666,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.09881480038166046,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 5235.2,
|
| 1253 |
+
"valid_targets_min": 879
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.6336,
|
| 1257 |
+
"grad_norm": 0.28813454432354996,
|
| 1258 |
+
"learning_rate": 8.548730598224646e-06,
|
| 1259 |
+
"loss": 0.3666,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.06468958407640457,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 3096.8,
|
| 1264 |
+
"valid_targets_min": 672
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.6656,
|
| 1268 |
+
"grad_norm": 0.2715471326270883,
|
| 1269 |
+
"learning_rate": 8.186769522352053e-06,
|
| 1270 |
+
"loss": 0.3537,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.12461410462856293,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 4765.1,
|
| 1275 |
+
"valid_targets_min": 1146
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.6976,
|
| 1279 |
+
"grad_norm": 0.24995429633219066,
|
| 1280 |
+
"learning_rate": 7.830656093012714e-06,
|
| 1281 |
+
"loss": 0.3741,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.08743943274021149,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 3921.1,
|
| 1286 |
+
"valid_targets_min": 1066
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.7296,
|
| 1290 |
+
"grad_norm": 0.2710951610884447,
|
| 1291 |
+
"learning_rate": 7.480566589291696e-06,
|
| 1292 |
+
"loss": 0.3494,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.10593526065349579,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 4516.6,
|
| 1297 |
+
"valid_targets_min": 745
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.7616,
|
| 1301 |
+
"grad_norm": 0.24554789964053841,
|
| 1302 |
+
"learning_rate": 7.1366743083812285e-06,
|
| 1303 |
+
"loss": 0.3526,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.07037960737943649,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 4179.7,
|
| 1308 |
+
"valid_targets_min": 471
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.7936,
|
| 1312 |
+
"grad_norm": 0.2740035607007115,
|
| 1313 |
+
"learning_rate": 6.799149479797101e-06,
|
| 1314 |
+
"loss": 0.3569,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.11404471844434738,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 4479.6,
|
| 1319 |
+
"valid_targets_min": 422
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.8256,
|
| 1323 |
+
"grad_norm": 0.27857174556705694,
|
| 1324 |
+
"learning_rate": 6.4681591811137e-06,
|
| 1325 |
+
"loss": 0.3638,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.11332307755947113,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 4525.1,
|
| 1330 |
+
"valid_targets_min": 1350
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.8576,
|
| 1334 |
+
"grad_norm": 0.27036674346841433,
|
| 1335 |
+
"learning_rate": 6.143867255259197e-06,
|
| 1336 |
+
"loss": 0.3624,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.09035474061965942,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 3910.2,
|
| 1341 |
+
"valid_targets_min": 789
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.8895999999999997,
|
| 1345 |
+
"grad_norm": 0.26713392574410405,
|
| 1346 |
+
"learning_rate": 5.8264342294119504e-06,
|
| 1347 |
+
"loss": 0.3476,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.07560218870639801,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 3616.6,
|
| 1352 |
+
"valid_targets_min": 685
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.9215999999999998,
|
| 1356 |
+
"grad_norm": 0.2551820724608469,
|
| 1357 |
+
"learning_rate": 5.516017235538258e-06,
|
| 1358 |
+
"loss": 0.348,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.06616517901420593,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 2984.4,
|
| 1363 |
+
"valid_targets_min": 545
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.9536,
|
| 1367 |
+
"grad_norm": 0.2805543365431347,
|
| 1368 |
+
"learning_rate": 5.212769932610695e-06,
|
| 1369 |
+
"loss": 0.3688,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.09327379614114761,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 3519.0,
|
| 1374 |
+
"valid_targets_min": 613
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.9856,
|
| 1378 |
+
"grad_norm": 0.26253382838902783,
|
| 1379 |
+
"learning_rate": 4.916842430545681e-06,
|
| 1380 |
+
"loss": 0.3563,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.08990518748760223,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 4271.6,
|
| 1385 |
+
"valid_targets_min": 957
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.0128,
|
| 1389 |
+
"grad_norm": 0.2752324527328725,
|
| 1390 |
+
"learning_rate": 4.628381215897837e-06,
|
| 1391 |
+
"loss": 0.3569,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.09592108428478241,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 4334.0,
|
| 1396 |
+
"valid_targets_min": 825
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.0448,
|
| 1400 |
+
"grad_norm": 0.28556223681750137,
|
| 1401 |
+
"learning_rate": 4.347529079347914e-06,
|
| 1402 |
+
"loss": 0.3666,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.08014878630638123,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 3147.9,
|
| 1407 |
+
"valid_targets_min": 371
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.0768,
|
| 1411 |
+
"grad_norm": 0.2500322876350698,
|
| 1412 |
+
"learning_rate": 4.074425045020247e-06,
|
| 1413 |
+
"loss": 0.344,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.09790447354316711,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 4746.9,
|
| 1418 |
+
"valid_targets_min": 831
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.1088,
|
| 1422 |
+
"grad_norm": 0.2658128147848985,
|
| 1423 |
+
"learning_rate": 3.8092043016646487e-06,
|
| 1424 |
+
"loss": 0.3557,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.11421743780374527,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 5214.5,
|
| 1429 |
+
"valid_targets_min": 760
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.1408,
|
| 1433 |
+
"grad_norm": 0.2620287255007106,
|
| 1434 |
+
"learning_rate": 3.551998135736867e-06,
|
| 1435 |
+
"loss": 0.3577,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.06395424902439117,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 3036.3,
|
| 1440 |
+
"valid_targets_min": 919
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.1728,
|
| 1444 |
+
"grad_norm": 0.27025685574336344,
|
| 1445 |
+
"learning_rate": 3.3029338664107267e-06,
|
| 1446 |
+
"loss": 0.3468,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.0907667949795723,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 4400.2,
|
| 1451 |
+
"valid_targets_min": 430
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.2048,
|
| 1455 |
+
"grad_norm": 0.25882052653428583,
|
| 1456 |
+
"learning_rate": 3.0621347825540625e-06,
|
| 1457 |
+
"loss": 0.3654,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.09003841876983643,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 4704.8,
|
| 1462 |
+
"valid_targets_min": 877
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.2368,
|
| 1466 |
+
"grad_norm": 0.30004947633921647,
|
| 1467 |
+
"learning_rate": 2.8297200816997183e-06,
|
| 1468 |
+
"loss": 0.3671,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.06004023551940918,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 2139.1,
|
| 1473 |
+
"valid_targets_min": 709
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.2688,
|
| 1477 |
+
"grad_norm": 0.26732020980047305,
|
| 1478 |
+
"learning_rate": 2.605804811041803e-06,
|
| 1479 |
+
"loss": 0.3547,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.09256140887737274,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 4849.9,
|
| 1484 |
+
"valid_targets_min": 958
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.3008,
|
| 1488 |
+
"grad_norm": 0.2590174884621007,
|
| 1489 |
+
"learning_rate": 2.390499810486351e-06,
|
| 1490 |
+
"loss": 0.3507,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.07109656929969788,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 3647.4,
|
| 1495 |
+
"valid_targets_min": 713
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.3328,
|
| 1499 |
+
"grad_norm": 0.2753475580823456,
|
| 1500 |
+
"learning_rate": 2.183911657784685e-06,
|
| 1501 |
+
"loss": 0.3456,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.09951435029506683,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 4019.5,
|
| 1506 |
+
"valid_targets_min": 738
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.3648,
|
| 1510 |
+
"grad_norm": 0.25668120929058674,
|
| 1511 |
+
"learning_rate": 1.986142615776532e-06,
|
| 1512 |
+
"loss": 0.3508,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.08455836772918701,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 3837.6,
|
| 1517 |
+
"valid_targets_min": 916
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.3968,
|
| 1521 |
+
"grad_norm": 0.7219968858898081,
|
| 1522 |
+
"learning_rate": 1.7972905817690644e-06,
|
| 1523 |
+
"loss": 0.3436,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.08186328411102295,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 2692.8,
|
| 1528 |
+
"valid_targets_min": 830
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.4288,
|
| 1532 |
+
"grad_norm": 0.2646298047140867,
|
| 1533 |
+
"learning_rate": 1.617449039076955e-06,
|
| 1534 |
+
"loss": 0.3422,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.06586353480815887,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 3125.6,
|
| 1539 |
+
"valid_targets_min": 779
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.4608,
|
| 1543 |
+
"grad_norm": 0.27486921576830325,
|
| 1544 |
+
"learning_rate": 1.4467070107473413e-06,
|
| 1545 |
+
"loss": 0.3557,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.10247328877449036,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 4226.4,
|
| 1550 |
+
"valid_targets_min": 872
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.4928,
|
| 1554 |
+
"grad_norm": 0.2521880662478794,
|
| 1555 |
+
"learning_rate": 1.2851490154926816e-06,
|
| 1556 |
+
"loss": 0.3593,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.08326166868209839,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 3732.3,
|
| 1561 |
+
"valid_targets_min": 745
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.5248,
|
| 1565 |
+
"grad_norm": 0.27665376943053505,
|
| 1566 |
+
"learning_rate": 1.1328550258533211e-06,
|
| 1567 |
+
"loss": 0.3519,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.11586222052574158,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 5429.9,
|
| 1572 |
+
"valid_targets_min": 659
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.5568,
|
| 1576 |
+
"grad_norm": 0.2873076430878492,
|
| 1577 |
+
"learning_rate": 9.899004286103953e-07,
|
| 1578 |
+
"loss": 0.3577,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.1280900090932846,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 4937.1,
|
| 1583 |
+
"valid_targets_min": 1150
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.5888,
|
| 1587 |
+
"grad_norm": 0.2613753606147518,
|
| 1588 |
+
"learning_rate": 8.5635598746876e-07,
|
| 1589 |
+
"loss": 0.3599,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.0687340646982193,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 3370.6,
|
| 1594 |
+
"valid_targets_min": 715
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.6208,
|
| 1598 |
+
"grad_norm": 0.23830838005131266,
|
| 1599 |
+
"learning_rate": 7.32287808028389e-07,
|
| 1600 |
+
"loss": 0.3432,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.09013132750988007,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 4541.8,
|
| 1605 |
+
"valid_targets_min": 954
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.6528,
|
| 1609 |
+
"grad_norm": 0.26521568774090987,
|
| 1610 |
+
"learning_rate": 6.177573050615327e-07,
|
| 1611 |
+
"loss": 0.348,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.05397915840148926,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 2744.1,
|
| 1616 |
+
"valid_targets_min": 523
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.6848,
|
| 1620 |
+
"grad_norm": 0.36669589815503967,
|
| 1621 |
+
"learning_rate": 5.128211721119213e-07,
|
| 1622 |
+
"loss": 0.3532,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.08455473929643631,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 3970.4,
|
| 1627 |
+
"valid_targets_min": 463
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.7168,
|
| 1631 |
+
"grad_norm": 0.2800304158375605,
|
| 1632 |
+
"learning_rate": 4.175313534309755e-07,
|
| 1633 |
+
"loss": 0.3537,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.055731166154146194,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 2201.8,
|
| 1638 |
+
"valid_targets_min": 653
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.7488,
|
| 1642 |
+
"grad_norm": 0.28968831896780245,
|
| 1643 |
+
"learning_rate": 3.319350182649861e-07,
|
| 1644 |
+
"loss": 0.358,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.10389368236064911,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 5040.3,
|
| 1649 |
+
"valid_targets_min": 676
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.7808,
|
| 1653 |
+
"grad_norm": 0.23644895830233334,
|
| 1654 |
+
"learning_rate": 2.560745375059392e-07,
|
| 1655 |
+
"loss": 0.3384,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.07365654408931732,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 4284.6,
|
| 1660 |
+
"valid_targets_min": 679
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.8128,
|
| 1664 |
+
"grad_norm": 0.2522036009533431,
|
| 1665 |
+
"learning_rate": 1.8998746271758016e-07,
|
| 1666 |
+
"loss": 0.3493,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.06476405262947083,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 3923.4,
|
| 1671 |
+
"valid_targets_min": 742
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.8448,
|
| 1675 |
+
"grad_norm": 0.2538354955682196,
|
| 1676 |
+
"learning_rate": 1.337065075470778e-07,
|
| 1677 |
+
"loss": 0.3465,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.08009777963161469,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 3415.3,
|
| 1682 |
+
"valid_targets_min": 784
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.8768,
|
| 1686 |
+
"grad_norm": 0.26003194613743286,
|
| 1687 |
+
"learning_rate": 8.725953153150279e-08,
|
| 1688 |
+
"loss": 0.3677,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.0702776163816452,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 3007.7,
|
| 1693 |
+
"valid_targets_min": 957
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.9088,
|
| 1697 |
+
"grad_norm": 0.2798975287738127,
|
| 1698 |
+
"learning_rate": 5.066952630711886e-08,
|
| 1699 |
+
"loss": 0.3516,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.10448385775089264,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 3724.1,
|
| 1704 |
+
"valid_targets_min": 898
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.9408,
|
| 1708 |
+
"grad_norm": 0.22457486103536708,
|
| 1709 |
+
"learning_rate": 2.3954604228342283e-08,
|
| 1710 |
+
"loss": 0.339,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.06149273365736008,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 3175.9,
|
| 1715 |
+
"valid_targets_min": 626
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 4.9728,
|
| 1719 |
+
"grad_norm": 0.2530148403115037,
|
| 1720 |
+
"learning_rate": 7.12798940197601e-09,
|
| 1721 |
+
"loss": 0.3584,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.07958637177944183,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 4460.8,
|
| 1726 |
+
"valid_targets_min": 615
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"grad_norm": 0.5731229517630574,
|
| 1731 |
+
"learning_rate": 1.9801114115480802e-10,
|
| 1732 |
+
"loss": 0.3488,
|
| 1733 |
+
"loss_nan_ranks": 0,
|
| 1734 |
+
"loss_rank_avg": 0.34793925285339355,
|
| 1735 |
+
"step": 785,
|
| 1736 |
+
"valid_targets_mean": 3301.4,
|
| 1737 |
+
"valid_targets_min": 867
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 5.0,
|
| 1741 |
+
"loss_nan_ranks": 0,
|
| 1742 |
+
"loss_rank_avg": 0.34793925285339355,
|
| 1743 |
+
"step": 785,
|
| 1744 |
+
"total_flos": 1.0929687562016522e+18,
|
| 1745 |
+
"train_loss": 0.3876832940775877,
|
| 1746 |
+
"train_runtime": 17345.5697,
|
| 1747 |
+
"train_samples_per_second": 2.883,
|
| 1748 |
+
"train_steps_per_second": 0.045,
|
| 1749 |
+
"valid_targets_mean": 3301.4,
|
| 1750 |
+
"valid_targets_min": 867
|
| 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": 1.0929687562016522e+18,
|
| 1771 |
+
"train_batch_size": 1,
|
| 1772 |
+
"trial_name": null,
|
| 1773 |
+
"trial_params": null
|
| 1774 |
+
}
|
training_loss.png
ADDED
|