End of training
Browse files- README.md +2 -1
- all_results.json +16 -0
- train_results.json +16 -0
- trainer_state.json +1763 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen3-8B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: all-puzzles-sandboxes-traces-terminus-2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# all-puzzles-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: all-puzzles-sandboxes-traces-terminus-2
|
|
|
|
| 16 |
|
| 17 |
# all-puzzles-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/all-puzzles-sandboxes-traces-terminus-2 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.226948326866271,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 203.27427376939005,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.23872976005077362,
|
| 7 |
+
"mfu_percent": 0.42739619078526503,
|
| 8 |
+
"mfu_percent_theoretical": 20.55351605352781,
|
| 9 |
+
"total_flos": 4.410167938263286e+17,
|
| 10 |
+
"train_loss": 0.3955344903163421,
|
| 11 |
+
"train_runtime": 13041.82,
|
| 12 |
+
"train_samples_per_second": 3.807,
|
| 13 |
+
"train_steps_per_second": 0.06,
|
| 14 |
+
"valid_targets_mean": 1017.1,
|
| 15 |
+
"valid_targets_min": 327
|
| 16 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"achieved_tflops_per_gpu": 4.226948326866271,
|
| 3 |
+
"achieved_tflops_per_gpu_theoretical": 203.27427376939005,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"loss_nan_ranks": 0,
|
| 6 |
+
"loss_rank_avg": 0.23872976005077362,
|
| 7 |
+
"mfu_percent": 0.42739619078526503,
|
| 8 |
+
"mfu_percent_theoretical": 20.55351605352781,
|
| 9 |
+
"total_flos": 4.410167938263286e+17,
|
| 10 |
+
"train_loss": 0.3955344903163421,
|
| 11 |
+
"train_runtime": 13041.82,
|
| 12 |
+
"train_samples_per_second": 3.807,
|
| 13 |
+
"train_steps_per_second": 0.06,
|
| 14 |
+
"valid_targets_mean": 1017.1,
|
| 15 |
+
"valid_targets_min": 327
|
| 16 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1763 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 780,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0322061191626409,
|
| 14 |
+
"grad_norm": 12.898349962866439,
|
| 15 |
+
"learning_rate": 2.0512820512820513e-06,
|
| 16 |
+
"loss": 0.8621,
|
| 17 |
+
"loss_nan_ranks": 0,
|
| 18 |
+
"loss_rank_avg": 0.11310747265815735,
|
| 19 |
+
"step": 5,
|
| 20 |
+
"valid_targets_mean": 552.5,
|
| 21 |
+
"valid_targets_min": 297
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.0644122383252818,
|
| 25 |
+
"grad_norm": 6.85206740052359,
|
| 26 |
+
"learning_rate": 4.615384615384616e-06,
|
| 27 |
+
"loss": 0.872,
|
| 28 |
+
"loss_nan_ranks": 0,
|
| 29 |
+
"loss_rank_avg": 0.11378927528858185,
|
| 30 |
+
"step": 10,
|
| 31 |
+
"valid_targets_mean": 1438.5,
|
| 32 |
+
"valid_targets_min": 332
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.0966183574879227,
|
| 36 |
+
"grad_norm": 3.405492909651102,
|
| 37 |
+
"learning_rate": 7.17948717948718e-06,
|
| 38 |
+
"loss": 0.7405,
|
| 39 |
+
"loss_nan_ranks": 0,
|
| 40 |
+
"loss_rank_avg": 0.06183416768908501,
|
| 41 |
+
"step": 15,
|
| 42 |
+
"valid_targets_mean": 580.8,
|
| 43 |
+
"valid_targets_min": 260
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.1288244766505636,
|
| 47 |
+
"grad_norm": 1.903925034713442,
|
| 48 |
+
"learning_rate": 9.743589743589744e-06,
|
| 49 |
+
"loss": 0.7208,
|
| 50 |
+
"loss_nan_ranks": 0,
|
| 51 |
+
"loss_rank_avg": 0.08290648460388184,
|
| 52 |
+
"step": 20,
|
| 53 |
+
"valid_targets_mean": 922.6,
|
| 54 |
+
"valid_targets_min": 337
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.1610305958132045,
|
| 58 |
+
"grad_norm": 1.3698491699971576,
|
| 59 |
+
"learning_rate": 1.230769230769231e-05,
|
| 60 |
+
"loss": 0.6653,
|
| 61 |
+
"loss_nan_ranks": 0,
|
| 62 |
+
"loss_rank_avg": 0.08934478461742401,
|
| 63 |
+
"step": 25,
|
| 64 |
+
"valid_targets_mean": 716.9,
|
| 65 |
+
"valid_targets_min": 355
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1932367149758454,
|
| 69 |
+
"grad_norm": 1.0986695170310006,
|
| 70 |
+
"learning_rate": 1.4871794871794874e-05,
|
| 71 |
+
"loss": 0.6276,
|
| 72 |
+
"loss_nan_ranks": 0,
|
| 73 |
+
"loss_rank_avg": 0.07862719893455505,
|
| 74 |
+
"step": 30,
|
| 75 |
+
"valid_targets_mean": 893.6,
|
| 76 |
+
"valid_targets_min": 342
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.22544283413848631,
|
| 80 |
+
"grad_norm": 0.8812267490478244,
|
| 81 |
+
"learning_rate": 1.7435897435897438e-05,
|
| 82 |
+
"loss": 0.615,
|
| 83 |
+
"loss_nan_ranks": 0,
|
| 84 |
+
"loss_rank_avg": 0.12738855183124542,
|
| 85 |
+
"step": 35,
|
| 86 |
+
"valid_targets_mean": 1057.4,
|
| 87 |
+
"valid_targets_min": 363
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2576489533011272,
|
| 91 |
+
"grad_norm": 0.8401277977595764,
|
| 92 |
+
"learning_rate": 2e-05,
|
| 93 |
+
"loss": 0.5848,
|
| 94 |
+
"loss_nan_ranks": 0,
|
| 95 |
+
"loss_rank_avg": 0.0551215335726738,
|
| 96 |
+
"step": 40,
|
| 97 |
+
"valid_targets_mean": 723.9,
|
| 98 |
+
"valid_targets_min": 292
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.2898550724637681,
|
| 102 |
+
"grad_norm": 0.6745283792399535,
|
| 103 |
+
"learning_rate": 2.2564102564102566e-05,
|
| 104 |
+
"loss": 0.5279,
|
| 105 |
+
"loss_nan_ranks": 0,
|
| 106 |
+
"loss_rank_avg": 0.05893056094646454,
|
| 107 |
+
"step": 45,
|
| 108 |
+
"valid_targets_mean": 787.1,
|
| 109 |
+
"valid_targets_min": 285
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.322061191626409,
|
| 113 |
+
"grad_norm": 0.6559777036049393,
|
| 114 |
+
"learning_rate": 2.512820512820513e-05,
|
| 115 |
+
"loss": 0.5298,
|
| 116 |
+
"loss_nan_ranks": 0,
|
| 117 |
+
"loss_rank_avg": 0.06752099096775055,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"valid_targets_mean": 897.5,
|
| 120 |
+
"valid_targets_min": 221
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.35426731078904994,
|
| 124 |
+
"grad_norm": 0.5512846197250572,
|
| 125 |
+
"learning_rate": 2.7692307692307694e-05,
|
| 126 |
+
"loss": 0.4835,
|
| 127 |
+
"loss_nan_ranks": 0,
|
| 128 |
+
"loss_rank_avg": 0.10814902186393738,
|
| 129 |
+
"step": 55,
|
| 130 |
+
"valid_targets_mean": 1876.0,
|
| 131 |
+
"valid_targets_min": 240
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.3864734299516908,
|
| 135 |
+
"grad_norm": 0.4993982742101459,
|
| 136 |
+
"learning_rate": 3.0256410256410257e-05,
|
| 137 |
+
"loss": 0.4953,
|
| 138 |
+
"loss_nan_ranks": 0,
|
| 139 |
+
"loss_rank_avg": 0.06639184057712555,
|
| 140 |
+
"step": 60,
|
| 141 |
+
"valid_targets_mean": 1391.9,
|
| 142 |
+
"valid_targets_min": 344
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.41867954911433175,
|
| 146 |
+
"grad_norm": 0.6383132839926805,
|
| 147 |
+
"learning_rate": 3.282051282051282e-05,
|
| 148 |
+
"loss": 0.4583,
|
| 149 |
+
"loss_nan_ranks": 0,
|
| 150 |
+
"loss_rank_avg": 0.07518483698368073,
|
| 151 |
+
"step": 65,
|
| 152 |
+
"valid_targets_mean": 898.4,
|
| 153 |
+
"valid_targets_min": 327
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.45088566827697263,
|
| 157 |
+
"grad_norm": 0.5621105173468922,
|
| 158 |
+
"learning_rate": 3.538461538461539e-05,
|
| 159 |
+
"loss": 0.469,
|
| 160 |
+
"loss_nan_ranks": 0,
|
| 161 |
+
"loss_rank_avg": 0.028787091374397278,
|
| 162 |
+
"step": 70,
|
| 163 |
+
"valid_targets_mean": 518.4,
|
| 164 |
+
"valid_targets_min": 304
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4830917874396135,
|
| 168 |
+
"grad_norm": 0.5544434814590523,
|
| 169 |
+
"learning_rate": 3.794871794871795e-05,
|
| 170 |
+
"loss": 0.4697,
|
| 171 |
+
"loss_nan_ranks": 0,
|
| 172 |
+
"loss_rank_avg": 0.04727187752723694,
|
| 173 |
+
"step": 75,
|
| 174 |
+
"valid_targets_mean": 674.9,
|
| 175 |
+
"valid_targets_min": 298
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.5152979066022544,
|
| 179 |
+
"grad_norm": 0.4882906127614797,
|
| 180 |
+
"learning_rate": 3.999979972589652e-05,
|
| 181 |
+
"loss": 0.4368,
|
| 182 |
+
"loss_nan_ranks": 0,
|
| 183 |
+
"loss_rank_avg": 0.04346778243780136,
|
| 184 |
+
"step": 80,
|
| 185 |
+
"valid_targets_mean": 946.8,
|
| 186 |
+
"valid_targets_min": 300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.5475040257648953,
|
| 190 |
+
"grad_norm": 0.9150926943645863,
|
| 191 |
+
"learning_rate": 3.999279055341771e-05,
|
| 192 |
+
"loss": 0.4631,
|
| 193 |
+
"loss_nan_ranks": 0,
|
| 194 |
+
"loss_rank_avg": 0.09739387035369873,
|
| 195 |
+
"step": 85,
|
| 196 |
+
"valid_targets_mean": 1282.0,
|
| 197 |
+
"valid_targets_min": 408
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.5797101449275363,
|
| 201 |
+
"grad_norm": 0.6709571649722427,
|
| 202 |
+
"learning_rate": 3.9975771686375434e-05,
|
| 203 |
+
"loss": 0.4388,
|
| 204 |
+
"loss_nan_ranks": 0,
|
| 205 |
+
"loss_rank_avg": 0.03802517428994179,
|
| 206 |
+
"step": 90,
|
| 207 |
+
"valid_targets_mean": 515.5,
|
| 208 |
+
"valid_targets_min": 329
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.6119162640901772,
|
| 212 |
+
"grad_norm": 0.8539827107022786,
|
| 213 |
+
"learning_rate": 3.9948751645524235e-05,
|
| 214 |
+
"loss": 0.4325,
|
| 215 |
+
"loss_nan_ranks": 0,
|
| 216 |
+
"loss_rank_avg": 0.062230631709098816,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"valid_targets_mean": 573.1,
|
| 219 |
+
"valid_targets_min": 335
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.644122383252818,
|
| 223 |
+
"grad_norm": 0.6365846252008591,
|
| 224 |
+
"learning_rate": 3.9911743958858376e-05,
|
| 225 |
+
"loss": 0.4566,
|
| 226 |
+
"loss_nan_ranks": 0,
|
| 227 |
+
"loss_rank_avg": 0.04755745083093643,
|
| 228 |
+
"step": 100,
|
| 229 |
+
"valid_targets_mean": 709.1,
|
| 230 |
+
"valid_targets_min": 322
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.6763285024154589,
|
| 234 |
+
"grad_norm": 0.4964193415485056,
|
| 235 |
+
"learning_rate": 3.9864767154838864e-05,
|
| 236 |
+
"loss": 0.4385,
|
| 237 |
+
"loss_nan_ranks": 0,
|
| 238 |
+
"loss_rank_avg": 0.05032774806022644,
|
| 239 |
+
"step": 105,
|
| 240 |
+
"valid_targets_mean": 1279.0,
|
| 241 |
+
"valid_targets_min": 295
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.7085346215780999,
|
| 245 |
+
"grad_norm": 0.5582500327494776,
|
| 246 |
+
"learning_rate": 3.9807844753116864e-05,
|
| 247 |
+
"loss": 0.4292,
|
| 248 |
+
"loss_nan_ranks": 0,
|
| 249 |
+
"loss_rank_avg": 0.13067539036273956,
|
| 250 |
+
"step": 110,
|
| 251 |
+
"valid_targets_mean": 2418.5,
|
| 252 |
+
"valid_targets_min": 276
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.7407407407407407,
|
| 256 |
+
"grad_norm": 0.5605455694371961,
|
| 257 |
+
"learning_rate": 3.9741005252758255e-05,
|
| 258 |
+
"loss": 0.4141,
|
| 259 |
+
"loss_nan_ranks": 0,
|
| 260 |
+
"loss_rank_avg": 0.03115726262331009,
|
| 261 |
+
"step": 115,
|
| 262 |
+
"valid_targets_mean": 681.2,
|
| 263 |
+
"valid_targets_min": 283
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.7729468599033816,
|
| 267 |
+
"grad_norm": 0.4737117451016544,
|
| 268 |
+
"learning_rate": 3.966428211797514e-05,
|
| 269 |
+
"loss": 0.441,
|
| 270 |
+
"loss_nan_ranks": 0,
|
| 271 |
+
"loss_rank_avg": 0.06667988002300262,
|
| 272 |
+
"step": 120,
|
| 273 |
+
"valid_targets_mean": 1534.2,
|
| 274 |
+
"valid_targets_min": 305
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.8051529790660226,
|
| 278 |
+
"grad_norm": 0.43538248655180545,
|
| 279 |
+
"learning_rate": 3.957771376137144e-05,
|
| 280 |
+
"loss": 0.4257,
|
| 281 |
+
"loss_nan_ranks": 0,
|
| 282 |
+
"loss_rank_avg": 0.04947737231850624,
|
| 283 |
+
"step": 125,
|
| 284 |
+
"valid_targets_mean": 1130.1,
|
| 285 |
+
"valid_targets_min": 308
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.8373590982286635,
|
| 289 |
+
"grad_norm": 0.48578578549548673,
|
| 290 |
+
"learning_rate": 3.9481343524711095e-05,
|
| 291 |
+
"loss": 0.4285,
|
| 292 |
+
"loss_nan_ranks": 0,
|
| 293 |
+
"loss_rank_avg": 0.036311663687229156,
|
| 294 |
+
"step": 130,
|
| 295 |
+
"valid_targets_mean": 1009.8,
|
| 296 |
+
"valid_targets_min": 333
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8695652173913043,
|
| 300 |
+
"grad_norm": 0.4857684316354408,
|
| 301 |
+
"learning_rate": 3.937521965721831e-05,
|
| 302 |
+
"loss": 0.4334,
|
| 303 |
+
"loss_nan_ranks": 0,
|
| 304 |
+
"loss_rank_avg": 0.058429233729839325,
|
| 305 |
+
"step": 135,
|
| 306 |
+
"valid_targets_mean": 1255.6,
|
| 307 |
+
"valid_targets_min": 379
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.9017713365539453,
|
| 311 |
+
"grad_norm": 0.536215147216915,
|
| 312 |
+
"learning_rate": 3.9259395291420864e-05,
|
| 313 |
+
"loss": 0.4324,
|
| 314 |
+
"loss_nan_ranks": 0,
|
| 315 |
+
"loss_rank_avg": 0.02372780255973339,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"valid_targets_mean": 394.0,
|
| 318 |
+
"valid_targets_min": 283
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.9339774557165862,
|
| 322 |
+
"grad_norm": 0.4911344506126531,
|
| 323 |
+
"learning_rate": 3.913392841654851e-05,
|
| 324 |
+
"loss": 0.4102,
|
| 325 |
+
"loss_nan_ranks": 0,
|
| 326 |
+
"loss_rank_avg": 0.021754739806056023,
|
| 327 |
+
"step": 145,
|
| 328 |
+
"valid_targets_mean": 400.6,
|
| 329 |
+
"valid_targets_min": 290
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.966183574879227,
|
| 333 |
+
"grad_norm": 0.5895368205854837,
|
| 334 |
+
"learning_rate": 3.8998881849499756e-05,
|
| 335 |
+
"loss": 0.4187,
|
| 336 |
+
"loss_nan_ranks": 0,
|
| 337 |
+
"loss_rank_avg": 0.10034637898206711,
|
| 338 |
+
"step": 150,
|
| 339 |
+
"valid_targets_mean": 1143.1,
|
| 340 |
+
"valid_targets_min": 321
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.998389694041868,
|
| 344 |
+
"grad_norm": 0.4934508890573737,
|
| 345 |
+
"learning_rate": 3.885432320339167e-05,
|
| 346 |
+
"loss": 0.4174,
|
| 347 |
+
"loss_nan_ranks": 0,
|
| 348 |
+
"loss_rank_avg": 0.08337146043777466,
|
| 349 |
+
"step": 155,
|
| 350 |
+
"valid_targets_mean": 2523.6,
|
| 351 |
+
"valid_targets_min": 281
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.0257648953301126,
|
| 355 |
+
"grad_norm": 0.6397313297428109,
|
| 356 |
+
"learning_rate": 3.8700324853708304e-05,
|
| 357 |
+
"loss": 0.4417,
|
| 358 |
+
"loss_nan_ranks": 0,
|
| 359 |
+
"loss_rank_avg": 0.059281207621097565,
|
| 360 |
+
"step": 160,
|
| 361 |
+
"valid_targets_mean": 753.4,
|
| 362 |
+
"valid_targets_min": 309
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.0579710144927537,
|
| 366 |
+
"grad_norm": 0.46415104272287283,
|
| 367 |
+
"learning_rate": 3.853696390206484e-05,
|
| 368 |
+
"loss": 0.3984,
|
| 369 |
+
"loss_nan_ranks": 0,
|
| 370 |
+
"loss_rank_avg": 0.028974976390600204,
|
| 371 |
+
"step": 165,
|
| 372 |
+
"valid_targets_mean": 518.6,
|
| 373 |
+
"valid_targets_min": 336
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0901771336553945,
|
| 377 |
+
"grad_norm": 0.5352714190562025,
|
| 378 |
+
"learning_rate": 3.8364322137605484e-05,
|
| 379 |
+
"loss": 0.3912,
|
| 380 |
+
"loss_nan_ranks": 0,
|
| 381 |
+
"loss_rank_avg": 0.054242536425590515,
|
| 382 |
+
"step": 170,
|
| 383 |
+
"valid_targets_mean": 1146.2,
|
| 384 |
+
"valid_targets_min": 307
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.1223832528180355,
|
| 388 |
+
"grad_norm": 0.5623597503666744,
|
| 389 |
+
"learning_rate": 3.818248599605448e-05,
|
| 390 |
+
"loss": 0.3861,
|
| 391 |
+
"loss_nan_ranks": 0,
|
| 392 |
+
"loss_rank_avg": 0.0729055255651474,
|
| 393 |
+
"step": 175,
|
| 394 |
+
"valid_targets_mean": 1848.2,
|
| 395 |
+
"valid_targets_min": 325
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1545893719806763,
|
| 399 |
+
"grad_norm": 0.5742605838509625,
|
| 400 |
+
"learning_rate": 3.799154651644075e-05,
|
| 401 |
+
"loss": 0.397,
|
| 402 |
+
"loss_nan_ranks": 0,
|
| 403 |
+
"loss_rank_avg": 0.05786103010177612,
|
| 404 |
+
"step": 180,
|
| 405 |
+
"valid_targets_mean": 768.2,
|
| 406 |
+
"valid_targets_min": 255
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.1867954911433172,
|
| 410 |
+
"grad_norm": 0.48599772051318924,
|
| 411 |
+
"learning_rate": 3.7791599295517825e-05,
|
| 412 |
+
"loss": 0.3868,
|
| 413 |
+
"loss_nan_ranks": 0,
|
| 414 |
+
"loss_rank_avg": 0.044465214014053345,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"valid_targets_mean": 953.5,
|
| 417 |
+
"valid_targets_min": 305
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2190016103059582,
|
| 421 |
+
"grad_norm": 0.4427239963256647,
|
| 422 |
+
"learning_rate": 3.75827444399019e-05,
|
| 423 |
+
"loss": 0.4163,
|
| 424 |
+
"loss_nan_ranks": 0,
|
| 425 |
+
"loss_rank_avg": 0.03635489195585251,
|
| 426 |
+
"step": 190,
|
| 427 |
+
"valid_targets_mean": 787.0,
|
| 428 |
+
"valid_targets_min": 287
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.251207729468599,
|
| 432 |
+
"grad_norm": 0.46071305423252523,
|
| 433 |
+
"learning_rate": 3.736508651595188e-05,
|
| 434 |
+
"loss": 0.3759,
|
| 435 |
+
"loss_nan_ranks": 0,
|
| 436 |
+
"loss_rank_avg": 0.1078839898109436,
|
| 437 |
+
"step": 195,
|
| 438 |
+
"valid_targets_mean": 1857.1,
|
| 439 |
+
"valid_targets_min": 310
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2834138486312399,
|
| 443 |
+
"grad_norm": 0.45110054102426617,
|
| 444 |
+
"learning_rate": 3.713873449741669e-05,
|
| 445 |
+
"loss": 0.4067,
|
| 446 |
+
"loss_nan_ranks": 0,
|
| 447 |
+
"loss_rank_avg": 0.040066909044981,
|
| 448 |
+
"step": 200,
|
| 449 |
+
"valid_targets_mean": 1158.4,
|
| 450 |
+
"valid_targets_min": 337
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.315619967793881,
|
| 454 |
+
"grad_norm": 0.4952144859012548,
|
| 455 |
+
"learning_rate": 3.69038017108759e-05,
|
| 456 |
+
"loss": 0.3931,
|
| 457 |
+
"loss_nan_ranks": 0,
|
| 458 |
+
"loss_rank_avg": 0.02214495651423931,
|
| 459 |
+
"step": 205,
|
| 460 |
+
"valid_targets_mean": 509.6,
|
| 461 |
+
"valid_targets_min": 257
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.3478260869565217,
|
| 465 |
+
"grad_norm": 0.4303160555876925,
|
| 466 |
+
"learning_rate": 3.6660405779001036e-05,
|
| 467 |
+
"loss": 0.4067,
|
| 468 |
+
"loss_nan_ranks": 0,
|
| 469 |
+
"loss_rank_avg": 0.10445412993431091,
|
| 470 |
+
"step": 210,
|
| 471 |
+
"valid_targets_mean": 3890.0,
|
| 472 |
+
"valid_targets_min": 318
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3800322061191626,
|
| 476 |
+
"grad_norm": 0.4365789322635305,
|
| 477 |
+
"learning_rate": 3.640866856166601e-05,
|
| 478 |
+
"loss": 0.3827,
|
| 479 |
+
"loss_nan_ranks": 0,
|
| 480 |
+
"loss_rank_avg": 0.0636882334947586,
|
| 481 |
+
"step": 215,
|
| 482 |
+
"valid_targets_mean": 1477.5,
|
| 483 |
+
"valid_targets_min": 390
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.4122383252818036,
|
| 487 |
+
"grad_norm": 0.4837420644157074,
|
| 488 |
+
"learning_rate": 3.614871609493614e-05,
|
| 489 |
+
"loss": 0.3846,
|
| 490 |
+
"loss_nan_ranks": 0,
|
| 491 |
+
"loss_rank_avg": 0.05718361958861351,
|
| 492 |
+
"step": 220,
|
| 493 |
+
"valid_targets_mean": 1256.9,
|
| 494 |
+
"valid_targets_min": 353
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4444444444444444,
|
| 498 |
+
"grad_norm": 0.47344211330731195,
|
| 499 |
+
"learning_rate": 3.5880678527966224e-05,
|
| 500 |
+
"loss": 0.3729,
|
| 501 |
+
"loss_nan_ranks": 0,
|
| 502 |
+
"loss_rank_avg": 0.07174763828516006,
|
| 503 |
+
"step": 225,
|
| 504 |
+
"valid_targets_mean": 1642.0,
|
| 505 |
+
"valid_targets_min": 260
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.4766505636070852,
|
| 509 |
+
"grad_norm": 0.5547472032191869,
|
| 510 |
+
"learning_rate": 3.5604690057839456e-05,
|
| 511 |
+
"loss": 0.3808,
|
| 512 |
+
"loss_nan_ranks": 0,
|
| 513 |
+
"loss_rank_avg": 0.04665002226829529,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"valid_targets_mean": 620.6,
|
| 516 |
+
"valid_targets_min": 358
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.5088566827697263,
|
| 520 |
+
"grad_norm": 0.5106632095827158,
|
| 521 |
+
"learning_rate": 3.532088886237956e-05,
|
| 522 |
+
"loss": 0.3796,
|
| 523 |
+
"loss_nan_ranks": 0,
|
| 524 |
+
"loss_rank_avg": 0.060500893741846085,
|
| 525 |
+
"step": 235,
|
| 526 |
+
"valid_targets_mean": 1519.4,
|
| 527 |
+
"valid_targets_min": 288
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.541062801932367,
|
| 531 |
+
"grad_norm": 0.4498419929162088,
|
| 532 |
+
"learning_rate": 3.5029417030970005e-05,
|
| 533 |
+
"loss": 0.4032,
|
| 534 |
+
"loss_nan_ranks": 0,
|
| 535 |
+
"loss_rank_avg": 0.027759376913309097,
|
| 536 |
+
"step": 240,
|
| 537 |
+
"valid_targets_mean": 679.4,
|
| 538 |
+
"valid_targets_min": 294
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.573268921095008,
|
| 542 |
+
"grad_norm": 0.4631774519072297,
|
| 543 |
+
"learning_rate": 3.473042049341474e-05,
|
| 544 |
+
"loss": 0.3899,
|
| 545 |
+
"loss_nan_ranks": 0,
|
| 546 |
+
"loss_rank_avg": 0.04057846963405609,
|
| 547 |
+
"step": 245,
|
| 548 |
+
"valid_targets_mean": 1025.0,
|
| 549 |
+
"valid_targets_min": 218
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.605475040257649,
|
| 553 |
+
"grad_norm": 0.47311421989321817,
|
| 554 |
+
"learning_rate": 3.442404894687629e-05,
|
| 555 |
+
"loss": 0.395,
|
| 556 |
+
"loss_nan_ranks": 0,
|
| 557 |
+
"loss_rank_avg": 0.027594510465860367,
|
| 558 |
+
"step": 250,
|
| 559 |
+
"valid_targets_mean": 574.0,
|
| 560 |
+
"valid_targets_min": 269
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.6376811594202898,
|
| 564 |
+
"grad_norm": 0.6220706207123248,
|
| 565 |
+
"learning_rate": 3.411045578092754e-05,
|
| 566 |
+
"loss": 0.4178,
|
| 567 |
+
"loss_nan_ranks": 0,
|
| 568 |
+
"loss_rank_avg": 0.02938500978052616,
|
| 569 |
+
"step": 255,
|
| 570 |
+
"valid_targets_mean": 496.4,
|
| 571 |
+
"valid_targets_min": 325
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.6698872785829306,
|
| 575 |
+
"grad_norm": 0.5013668028351838,
|
| 576 |
+
"learning_rate": 3.3789798000754924e-05,
|
| 577 |
+
"loss": 0.3814,
|
| 578 |
+
"loss_nan_ranks": 0,
|
| 579 |
+
"loss_rank_avg": 0.027927782386541367,
|
| 580 |
+
"step": 260,
|
| 581 |
+
"valid_targets_mean": 767.4,
|
| 582 |
+
"valid_targets_min": 263
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.7020933977455717,
|
| 586 |
+
"grad_norm": 0.5397613733002328,
|
| 587 |
+
"learning_rate": 3.34622361485514e-05,
|
| 588 |
+
"loss": 0.3818,
|
| 589 |
+
"loss_nan_ranks": 0,
|
| 590 |
+
"loss_rank_avg": 0.03136827051639557,
|
| 591 |
+
"step": 265,
|
| 592 |
+
"valid_targets_mean": 570.8,
|
| 593 |
+
"valid_targets_min": 341
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7342995169082127,
|
| 597 |
+
"grad_norm": 0.5549058173437013,
|
| 598 |
+
"learning_rate": 3.31279342231386e-05,
|
| 599 |
+
"loss": 0.3931,
|
| 600 |
+
"loss_nan_ranks": 0,
|
| 601 |
+
"loss_rank_avg": 0.05748458951711655,
|
| 602 |
+
"step": 270,
|
| 603 |
+
"valid_targets_mean": 809.4,
|
| 604 |
+
"valid_targets_min": 263
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7665056360708533,
|
| 608 |
+
"grad_norm": 0.5121030877701617,
|
| 609 |
+
"learning_rate": 3.278705959785821e-05,
|
| 610 |
+
"loss": 0.4033,
|
| 611 |
+
"loss_nan_ranks": 0,
|
| 612 |
+
"loss_rank_avg": 0.03404642269015312,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"valid_targets_mean": 479.6,
|
| 615 |
+
"valid_targets_min": 275
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.7987117552334944,
|
| 619 |
+
"grad_norm": 0.4382008464023779,
|
| 620 |
+
"learning_rate": 3.243978293677408e-05,
|
| 621 |
+
"loss": 0.3693,
|
| 622 |
+
"loss_nan_ranks": 0,
|
| 623 |
+
"loss_rank_avg": 0.03826048597693443,
|
| 624 |
+
"step": 280,
|
| 625 |
+
"valid_targets_mean": 1542.5,
|
| 626 |
+
"valid_targets_min": 317
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8309178743961354,
|
| 630 |
+
"grad_norm": 0.6050902730329194,
|
| 631 |
+
"learning_rate": 3.208627810922665e-05,
|
| 632 |
+
"loss": 0.3794,
|
| 633 |
+
"loss_nan_ranks": 0,
|
| 634 |
+
"loss_rank_avg": 0.058419372886419296,
|
| 635 |
+
"step": 285,
|
| 636 |
+
"valid_targets_mean": 855.5,
|
| 637 |
+
"valid_targets_min": 303
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.863123993558776,
|
| 641 |
+
"grad_norm": 0.5038584214895873,
|
| 642 |
+
"learning_rate": 3.172672210278254e-05,
|
| 643 |
+
"loss": 0.3898,
|
| 644 |
+
"loss_nan_ranks": 0,
|
| 645 |
+
"loss_rank_avg": 0.05743902176618576,
|
| 646 |
+
"step": 290,
|
| 647 |
+
"valid_targets_mean": 1938.6,
|
| 648 |
+
"valid_targets_min": 333
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.895330112721417,
|
| 652 |
+
"grad_norm": 0.564278397643887,
|
| 653 |
+
"learning_rate": 3.136129493462312e-05,
|
| 654 |
+
"loss": 0.4063,
|
| 655 |
+
"loss_nan_ranks": 0,
|
| 656 |
+
"loss_rank_avg": 0.04776623472571373,
|
| 657 |
+
"step": 295,
|
| 658 |
+
"valid_targets_mean": 636.6,
|
| 659 |
+
"valid_targets_min": 333
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.927536231884058,
|
| 663 |
+
"grad_norm": 0.5937603264774977,
|
| 664 |
+
"learning_rate": 3.0990179561416124e-05,
|
| 665 |
+
"loss": 0.3934,
|
| 666 |
+
"loss_nan_ranks": 0,
|
| 667 |
+
"loss_rank_avg": 0.06801211833953857,
|
| 668 |
+
"step": 300,
|
| 669 |
+
"valid_targets_mean": 944.6,
|
| 670 |
+
"valid_targets_min": 340
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.959742351046699,
|
| 674 |
+
"grad_norm": 0.4964999875942384,
|
| 675 |
+
"learning_rate": 3.061356178771564e-05,
|
| 676 |
+
"loss": 0.3916,
|
| 677 |
+
"loss_nan_ranks": 0,
|
| 678 |
+
"loss_rank_avg": 0.03988070785999298,
|
| 679 |
+
"step": 305,
|
| 680 |
+
"valid_targets_mean": 664.6,
|
| 681 |
+
"valid_targets_min": 292
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.9919484702093397,
|
| 685 |
+
"grad_norm": 0.5661676600187325,
|
| 686 |
+
"learning_rate": 3.0231630172936205e-05,
|
| 687 |
+
"loss": 0.397,
|
| 688 |
+
"loss_nan_ranks": 0,
|
| 689 |
+
"loss_rank_avg": 0.041811276227235794,
|
| 690 |
+
"step": 310,
|
| 691 |
+
"valid_targets_mean": 593.0,
|
| 692 |
+
"valid_targets_min": 265
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.0193236714975846,
|
| 696 |
+
"grad_norm": 0.5035992902460946,
|
| 697 |
+
"learning_rate": 2.98445759369477e-05,
|
| 698 |
+
"loss": 0.3887,
|
| 699 |
+
"loss_nan_ranks": 0,
|
| 700 |
+
"loss_rank_avg": 0.024304592981934547,
|
| 701 |
+
"step": 315,
|
| 702 |
+
"valid_targets_mean": 545.0,
|
| 703 |
+
"valid_targets_min": 335
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0515297906602252,
|
| 707 |
+
"grad_norm": 0.4746849218727487,
|
| 708 |
+
"learning_rate": 2.945259286433819e-05,
|
| 709 |
+
"loss": 0.3821,
|
| 710 |
+
"loss_nan_ranks": 0,
|
| 711 |
+
"loss_rank_avg": 0.035836152732372284,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"valid_targets_mean": 911.9,
|
| 714 |
+
"valid_targets_min": 302
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.0837359098228663,
|
| 718 |
+
"grad_norm": 0.6069458438713761,
|
| 719 |
+
"learning_rate": 2.9055877207392752e-05,
|
| 720 |
+
"loss": 0.3795,
|
| 721 |
+
"loss_nan_ranks": 0,
|
| 722 |
+
"loss_rank_avg": 0.04563501104712486,
|
| 723 |
+
"step": 325,
|
| 724 |
+
"valid_targets_mean": 767.9,
|
| 725 |
+
"valid_targets_min": 316
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.1159420289855073,
|
| 729 |
+
"grad_norm": 0.464114101694879,
|
| 730 |
+
"learning_rate": 2.8654627587836724e-05,
|
| 731 |
+
"loss": 0.3588,
|
| 732 |
+
"loss_nan_ranks": 0,
|
| 733 |
+
"loss_rank_avg": 0.0893658697605133,
|
| 734 |
+
"step": 330,
|
| 735 |
+
"valid_targets_mean": 2450.0,
|
| 736 |
+
"valid_targets_min": 416
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.148148148148148,
|
| 740 |
+
"grad_norm": 0.4979196420820879,
|
| 741 |
+
"learning_rate": 2.8249044897392814e-05,
|
| 742 |
+
"loss": 0.3649,
|
| 743 |
+
"loss_nan_ranks": 0,
|
| 744 |
+
"loss_rank_avg": 0.051502641290426254,
|
| 745 |
+
"step": 335,
|
| 746 |
+
"valid_targets_mean": 1717.1,
|
| 747 |
+
"valid_targets_min": 244
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.180354267310789,
|
| 751 |
+
"grad_norm": 0.5588899086430609,
|
| 752 |
+
"learning_rate": 2.7839332197201505e-05,
|
| 753 |
+
"loss": 0.3799,
|
| 754 |
+
"loss_nan_ranks": 0,
|
| 755 |
+
"loss_rank_avg": 0.045202601701021194,
|
| 756 |
+
"step": 340,
|
| 757 |
+
"valid_targets_mean": 951.6,
|
| 758 |
+
"valid_targets_min": 360
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.21256038647343,
|
| 762 |
+
"grad_norm": 0.4764396199562437,
|
| 763 |
+
"learning_rate": 2.7425694616155474e-05,
|
| 764 |
+
"loss": 0.3623,
|
| 765 |
+
"loss_nan_ranks": 0,
|
| 766 |
+
"loss_rank_avg": 0.06349411606788635,
|
| 767 |
+
"step": 345,
|
| 768 |
+
"valid_targets_mean": 1610.1,
|
| 769 |
+
"valid_targets_min": 298
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.244766505636071,
|
| 773 |
+
"grad_norm": 0.4543103599849946,
|
| 774 |
+
"learning_rate": 2.700833924819868e-05,
|
| 775 |
+
"loss": 0.3587,
|
| 776 |
+
"loss_nan_ranks": 0,
|
| 777 |
+
"loss_rank_avg": 0.042356789112091064,
|
| 778 |
+
"step": 350,
|
| 779 |
+
"valid_targets_mean": 1044.9,
|
| 780 |
+
"valid_targets_min": 348
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.2769726247987117,
|
| 784 |
+
"grad_norm": 0.4714348506524609,
|
| 785 |
+
"learning_rate": 2.6587475048641596e-05,
|
| 786 |
+
"loss": 0.3681,
|
| 787 |
+
"loss_nan_ranks": 0,
|
| 788 |
+
"loss_rank_avg": 0.040921151638031006,
|
| 789 |
+
"step": 355,
|
| 790 |
+
"valid_targets_mean": 958.1,
|
| 791 |
+
"valid_targets_min": 252
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.3091787439613527,
|
| 795 |
+
"grad_norm": 0.45052091460255717,
|
| 796 |
+
"learning_rate": 2.6163312729544616e-05,
|
| 797 |
+
"loss": 0.3651,
|
| 798 |
+
"loss_nan_ranks": 0,
|
| 799 |
+
"loss_rank_avg": 0.0605086013674736,
|
| 800 |
+
"step": 360,
|
| 801 |
+
"valid_targets_mean": 1398.6,
|
| 802 |
+
"valid_targets_min": 407
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.3413848631239937,
|
| 806 |
+
"grad_norm": 0.8153337300604954,
|
| 807 |
+
"learning_rate": 2.5736064654221808e-05,
|
| 808 |
+
"loss": 0.3775,
|
| 809 |
+
"loss_nan_ranks": 0,
|
| 810 |
+
"loss_rank_avg": 0.035735368728637695,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"valid_targets_mean": 977.4,
|
| 813 |
+
"valid_targets_min": 308
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 2.3735909822866343,
|
| 817 |
+
"grad_norm": 0.469458881165414,
|
| 818 |
+
"learning_rate": 2.5305944730918017e-05,
|
| 819 |
+
"loss": 0.3979,
|
| 820 |
+
"loss_nan_ranks": 0,
|
| 821 |
+
"loss_rank_avg": 0.031210483983159065,
|
| 822 |
+
"step": 370,
|
| 823 |
+
"valid_targets_mean": 778.0,
|
| 824 |
+
"valid_targets_min": 251
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.4057971014492754,
|
| 828 |
+
"grad_norm": 0.5033946973514036,
|
| 829 |
+
"learning_rate": 2.487316830571244e-05,
|
| 830 |
+
"loss": 0.344,
|
| 831 |
+
"loss_nan_ranks": 0,
|
| 832 |
+
"loss_rank_avg": 0.0375363752245903,
|
| 833 |
+
"step": 375,
|
| 834 |
+
"valid_targets_mean": 912.9,
|
| 835 |
+
"valid_targets_min": 282
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4380032206119164,
|
| 839 |
+
"grad_norm": 0.5816567624854461,
|
| 840 |
+
"learning_rate": 2.4437952054702354e-05,
|
| 841 |
+
"loss": 0.3515,
|
| 842 |
+
"loss_nan_ranks": 0,
|
| 843 |
+
"loss_rank_avg": 0.029283087700605392,
|
| 844 |
+
"step": 380,
|
| 845 |
+
"valid_targets_mean": 378.5,
|
| 846 |
+
"valid_targets_min": 253
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.470209339774557,
|
| 850 |
+
"grad_norm": 0.43089304741489437,
|
| 851 |
+
"learning_rate": 2.4000513875520892e-05,
|
| 852 |
+
"loss": 0.349,
|
| 853 |
+
"loss_nan_ranks": 0,
|
| 854 |
+
"loss_rank_avg": 0.05593574792146683,
|
| 855 |
+
"step": 385,
|
| 856 |
+
"valid_targets_mean": 2434.5,
|
| 857 |
+
"valid_targets_min": 326
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.502415458937198,
|
| 861 |
+
"grad_norm": 0.45716732308341723,
|
| 862 |
+
"learning_rate": 2.3561072778243335e-05,
|
| 863 |
+
"loss": 0.3504,
|
| 864 |
+
"loss_nan_ranks": 0,
|
| 865 |
+
"loss_rank_avg": 0.033897001296281815,
|
| 866 |
+
"step": 390,
|
| 867 |
+
"valid_targets_mean": 945.1,
|
| 868 |
+
"valid_targets_min": 355
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.534621578099839,
|
| 872 |
+
"grad_norm": 0.5807110440832902,
|
| 873 |
+
"learning_rate": 2.311984877573636e-05,
|
| 874 |
+
"loss": 0.3755,
|
| 875 |
+
"loss_nan_ranks": 0,
|
| 876 |
+
"loss_rank_avg": 0.050817061215639114,
|
| 877 |
+
"step": 395,
|
| 878 |
+
"valid_targets_mean": 629.4,
|
| 879 |
+
"valid_targets_min": 297
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.5668276972624797,
|
| 883 |
+
"grad_norm": 0.48048954818776785,
|
| 884 |
+
"learning_rate": 2.2677062773505235e-05,
|
| 885 |
+
"loss": 0.3542,
|
| 886 |
+
"loss_nan_ranks": 0,
|
| 887 |
+
"loss_rank_avg": 0.026756778359413147,
|
| 888 |
+
"step": 400,
|
| 889 |
+
"valid_targets_mean": 713.8,
|
| 890 |
+
"valid_targets_min": 343
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.5990338164251208,
|
| 894 |
+
"grad_norm": 0.571474727405292,
|
| 895 |
+
"learning_rate": 2.2232936459094158e-05,
|
| 896 |
+
"loss": 0.3498,
|
| 897 |
+
"loss_nan_ranks": 0,
|
| 898 |
+
"loss_rank_avg": 0.03278893977403641,
|
| 899 |
+
"step": 405,
|
| 900 |
+
"valid_targets_mean": 543.1,
|
| 901 |
+
"valid_targets_min": 348
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.631239935587762,
|
| 905 |
+
"grad_norm": 0.46213821153329887,
|
| 906 |
+
"learning_rate": 2.1787692191094986e-05,
|
| 907 |
+
"loss": 0.3661,
|
| 908 |
+
"loss_nan_ranks": 0,
|
| 909 |
+
"loss_rank_avg": 0.02265809290111065,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"valid_targets_mean": 691.4,
|
| 912 |
+
"valid_targets_min": 264
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.6634460547504024,
|
| 916 |
+
"grad_norm": 0.4201064077434344,
|
| 917 |
+
"learning_rate": 2.1341552887820048e-05,
|
| 918 |
+
"loss": 0.3667,
|
| 919 |
+
"loss_nan_ranks": 0,
|
| 920 |
+
"loss_rank_avg": 0.04259542003273964,
|
| 921 |
+
"step": 415,
|
| 922 |
+
"valid_targets_mean": 1292.5,
|
| 923 |
+
"valid_targets_min": 356
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.6956521739130435,
|
| 927 |
+
"grad_norm": 0.5548360913759484,
|
| 928 |
+
"learning_rate": 2.0894741915694685e-05,
|
| 929 |
+
"loss": 0.3912,
|
| 930 |
+
"loss_nan_ranks": 0,
|
| 931 |
+
"loss_rank_avg": 0.04088819772005081,
|
| 932 |
+
"step": 420,
|
| 933 |
+
"valid_targets_mean": 663.2,
|
| 934 |
+
"valid_targets_min": 347
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.7278582930756845,
|
| 938 |
+
"grad_norm": 0.5304481578098202,
|
| 939 |
+
"learning_rate": 2.0447482977425465e-05,
|
| 940 |
+
"loss": 0.3596,
|
| 941 |
+
"loss_nan_ranks": 0,
|
| 942 |
+
"loss_rank_avg": 0.04688024893403053,
|
| 943 |
+
"step": 425,
|
| 944 |
+
"valid_targets_mean": 819.1,
|
| 945 |
+
"valid_targets_min": 324
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.760064412238325,
|
| 949 |
+
"grad_norm": 0.5165835397260369,
|
| 950 |
+
"learning_rate": 2e-05,
|
| 951 |
+
"loss": 0.3846,
|
| 952 |
+
"loss_nan_ranks": 0,
|
| 953 |
+
"loss_rank_avg": 0.028544938191771507,
|
| 954 |
+
"step": 430,
|
| 955 |
+
"valid_targets_mean": 738.5,
|
| 956 |
+
"valid_targets_min": 337
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.792270531400966,
|
| 960 |
+
"grad_norm": 0.4492097572019024,
|
| 961 |
+
"learning_rate": 1.9552517022574542e-05,
|
| 962 |
+
"loss": 0.3614,
|
| 963 |
+
"loss_nan_ranks": 0,
|
| 964 |
+
"loss_rank_avg": 0.03934304788708687,
|
| 965 |
+
"step": 435,
|
| 966 |
+
"valid_targets_mean": 950.9,
|
| 967 |
+
"valid_targets_min": 317
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 2.824476650563607,
|
| 971 |
+
"grad_norm": 0.5323764358603231,
|
| 972 |
+
"learning_rate": 1.910525808430532e-05,
|
| 973 |
+
"loss": 0.3961,
|
| 974 |
+
"loss_nan_ranks": 0,
|
| 975 |
+
"loss_rank_avg": 0.041988566517829895,
|
| 976 |
+
"step": 440,
|
| 977 |
+
"valid_targets_mean": 720.8,
|
| 978 |
+
"valid_targets_min": 321
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8566827697262482,
|
| 982 |
+
"grad_norm": 0.461336682212016,
|
| 983 |
+
"learning_rate": 1.8658447112179952e-05,
|
| 984 |
+
"loss": 0.3608,
|
| 985 |
+
"loss_nan_ranks": 0,
|
| 986 |
+
"loss_rank_avg": 0.039135366678237915,
|
| 987 |
+
"step": 445,
|
| 988 |
+
"valid_targets_mean": 938.9,
|
| 989 |
+
"valid_targets_min": 258
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.888888888888889,
|
| 993 |
+
"grad_norm": 0.3990312478235952,
|
| 994 |
+
"learning_rate": 1.821230780890502e-05,
|
| 995 |
+
"loss": 0.3463,
|
| 996 |
+
"loss_nan_ranks": 0,
|
| 997 |
+
"loss_rank_avg": 0.0437677726149559,
|
| 998 |
+
"step": 450,
|
| 999 |
+
"valid_targets_mean": 1247.5,
|
| 1000 |
+
"valid_targets_min": 344
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.92109500805153,
|
| 1004 |
+
"grad_norm": 0.549057571424804,
|
| 1005 |
+
"learning_rate": 1.776706354090585e-05,
|
| 1006 |
+
"loss": 0.3843,
|
| 1007 |
+
"loss_nan_ranks": 0,
|
| 1008 |
+
"loss_rank_avg": 0.06449249386787415,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"valid_targets_mean": 1010.9,
|
| 1011 |
+
"valid_targets_min": 296
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.9533011272141705,
|
| 1015 |
+
"grad_norm": 0.5774462386833141,
|
| 1016 |
+
"learning_rate": 1.7322937226494775e-05,
|
| 1017 |
+
"loss": 0.3701,
|
| 1018 |
+
"loss_nan_ranks": 0,
|
| 1019 |
+
"loss_rank_avg": 0.038682036101818085,
|
| 1020 |
+
"step": 460,
|
| 1021 |
+
"valid_targets_mean": 597.2,
|
| 1022 |
+
"valid_targets_min": 275
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.9855072463768115,
|
| 1026 |
+
"grad_norm": 0.45901296564550076,
|
| 1027 |
+
"learning_rate": 1.6880151224263646e-05,
|
| 1028 |
+
"loss": 0.3596,
|
| 1029 |
+
"loss_nan_ranks": 0,
|
| 1030 |
+
"loss_rank_avg": 0.04841262102127075,
|
| 1031 |
+
"step": 465,
|
| 1032 |
+
"valid_targets_mean": 1967.9,
|
| 1033 |
+
"valid_targets_min": 300
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 3.0128824476650564,
|
| 1037 |
+
"grad_norm": 0.5272617877932585,
|
| 1038 |
+
"learning_rate": 1.643892722175667e-05,
|
| 1039 |
+
"loss": 0.3841,
|
| 1040 |
+
"loss_nan_ranks": 0,
|
| 1041 |
+
"loss_rank_avg": 0.034066297113895416,
|
| 1042 |
+
"step": 470,
|
| 1043 |
+
"valid_targets_mean": 499.1,
|
| 1044 |
+
"valid_targets_min": 291
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0450885668276975,
|
| 1048 |
+
"grad_norm": 0.5673083616980614,
|
| 1049 |
+
"learning_rate": 1.5999486124479115e-05,
|
| 1050 |
+
"loss": 0.3609,
|
| 1051 |
+
"loss_nan_ranks": 0,
|
| 1052 |
+
"loss_rank_avg": 0.039988331496715546,
|
| 1053 |
+
"step": 475,
|
| 1054 |
+
"valid_targets_mean": 589.9,
|
| 1055 |
+
"valid_targets_min": 285
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.077294685990338,
|
| 1059 |
+
"grad_norm": 0.6094441969597235,
|
| 1060 |
+
"learning_rate": 1.5562047945297652e-05,
|
| 1061 |
+
"loss": 0.3685,
|
| 1062 |
+
"loss_nan_ranks": 0,
|
| 1063 |
+
"loss_rank_avg": 0.040959179401397705,
|
| 1064 |
+
"step": 480,
|
| 1065 |
+
"valid_targets_mean": 667.4,
|
| 1066 |
+
"valid_targets_min": 304
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.109500805152979,
|
| 1070 |
+
"grad_norm": 0.4240021705550922,
|
| 1071 |
+
"learning_rate": 1.5126831694287564e-05,
|
| 1072 |
+
"loss": 0.357,
|
| 1073 |
+
"loss_nan_ranks": 0,
|
| 1074 |
+
"loss_rank_avg": 0.026873812079429626,
|
| 1075 |
+
"step": 485,
|
| 1076 |
+
"valid_targets_mean": 763.5,
|
| 1077 |
+
"valid_targets_min": 303
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.14170692431562,
|
| 1081 |
+
"grad_norm": 0.46197206380035316,
|
| 1082 |
+
"learning_rate": 1.4694055269081987e-05,
|
| 1083 |
+
"loss": 0.3592,
|
| 1084 |
+
"loss_nan_ranks": 0,
|
| 1085 |
+
"loss_rank_avg": 0.028340086340904236,
|
| 1086 |
+
"step": 490,
|
| 1087 |
+
"valid_targets_mean": 740.5,
|
| 1088 |
+
"valid_targets_min": 305
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1739130434782608,
|
| 1092 |
+
"grad_norm": 0.443735264461422,
|
| 1093 |
+
"learning_rate": 1.4263935345778202e-05,
|
| 1094 |
+
"loss": 0.3604,
|
| 1095 |
+
"loss_nan_ranks": 0,
|
| 1096 |
+
"loss_rank_avg": 0.06339392066001892,
|
| 1097 |
+
"step": 495,
|
| 1098 |
+
"valid_targets_mean": 1977.4,
|
| 1099 |
+
"valid_targets_min": 302
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 3.206119162640902,
|
| 1103 |
+
"grad_norm": 0.44481778304862124,
|
| 1104 |
+
"learning_rate": 1.3836687270455388e-05,
|
| 1105 |
+
"loss": 0.3501,
|
| 1106 |
+
"loss_nan_ranks": 0,
|
| 1107 |
+
"loss_rank_avg": 0.027158450335264206,
|
| 1108 |
+
"step": 500,
|
| 1109 |
+
"valid_targets_mean": 723.2,
|
| 1110 |
+
"valid_targets_min": 253
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 3.238325281803543,
|
| 1114 |
+
"grad_norm": 0.4835785420746217,
|
| 1115 |
+
"learning_rate": 1.341252495135841e-05,
|
| 1116 |
+
"loss": 0.3512,
|
| 1117 |
+
"loss_nan_ranks": 0,
|
| 1118 |
+
"loss_rank_avg": 0.06052570044994354,
|
| 1119 |
+
"step": 505,
|
| 1120 |
+
"valid_targets_mean": 1464.1,
|
| 1121 |
+
"valid_targets_min": 244
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.2705314009661834,
|
| 1125 |
+
"grad_norm": 0.565332271009312,
|
| 1126 |
+
"learning_rate": 1.299166075180133e-05,
|
| 1127 |
+
"loss": 0.3572,
|
| 1128 |
+
"loss_nan_ranks": 0,
|
| 1129 |
+
"loss_rank_avg": 0.040864065289497375,
|
| 1130 |
+
"step": 510,
|
| 1131 |
+
"valid_targets_mean": 729.4,
|
| 1132 |
+
"valid_targets_min": 287
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.3027375201288245,
|
| 1136 |
+
"grad_norm": 0.4361579841431306,
|
| 1137 |
+
"learning_rate": 1.2574305383844528e-05,
|
| 1138 |
+
"loss": 0.3674,
|
| 1139 |
+
"loss_nan_ranks": 0,
|
| 1140 |
+
"loss_rank_avg": 0.05954238772392273,
|
| 1141 |
+
"step": 515,
|
| 1142 |
+
"valid_targets_mean": 2146.2,
|
| 1143 |
+
"valid_targets_min": 278
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3349436392914655,
|
| 1147 |
+
"grad_norm": 0.6338988451640348,
|
| 1148 |
+
"learning_rate": 1.2160667802798503e-05,
|
| 1149 |
+
"loss": 0.3485,
|
| 1150 |
+
"loss_nan_ranks": 0,
|
| 1151 |
+
"loss_rank_avg": 0.09376896917819977,
|
| 1152 |
+
"step": 520,
|
| 1153 |
+
"valid_targets_mean": 1447.6,
|
| 1154 |
+
"valid_targets_min": 249
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.367149758454106,
|
| 1158 |
+
"grad_norm": 0.4967736013730134,
|
| 1159 |
+
"learning_rate": 1.1750955102607193e-05,
|
| 1160 |
+
"loss": 0.346,
|
| 1161 |
+
"loss_nan_ranks": 0,
|
| 1162 |
+
"loss_rank_avg": 0.04483143985271454,
|
| 1163 |
+
"step": 525,
|
| 1164 |
+
"valid_targets_mean": 1566.4,
|
| 1165 |
+
"valid_targets_min": 313
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.399355877616747,
|
| 1169 |
+
"grad_norm": 0.3662991770351534,
|
| 1170 |
+
"learning_rate": 1.1345372412163276e-05,
|
| 1171 |
+
"loss": 0.3406,
|
| 1172 |
+
"loss_nan_ranks": 0,
|
| 1173 |
+
"loss_rank_avg": 0.02428291365504265,
|
| 1174 |
+
"step": 530,
|
| 1175 |
+
"valid_targets_mean": 1073.8,
|
| 1176 |
+
"valid_targets_min": 314
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.4315619967793882,
|
| 1180 |
+
"grad_norm": 0.5695193572731221,
|
| 1181 |
+
"learning_rate": 1.094412279260726e-05,
|
| 1182 |
+
"loss": 0.349,
|
| 1183 |
+
"loss_nan_ranks": 0,
|
| 1184 |
+
"loss_rank_avg": 0.03791878744959831,
|
| 1185 |
+
"step": 535,
|
| 1186 |
+
"valid_targets_mean": 812.0,
|
| 1187 |
+
"valid_targets_min": 293
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 3.463768115942029,
|
| 1191 |
+
"grad_norm": 0.456955054097705,
|
| 1192 |
+
"learning_rate": 1.0547407135661816e-05,
|
| 1193 |
+
"loss": 0.3597,
|
| 1194 |
+
"loss_nan_ranks": 0,
|
| 1195 |
+
"loss_rank_avg": 0.037714216858148575,
|
| 1196 |
+
"step": 540,
|
| 1197 |
+
"valid_targets_mean": 1002.9,
|
| 1198 |
+
"valid_targets_min": 312
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.49597423510467,
|
| 1202 |
+
"grad_norm": 0.5084684377157883,
|
| 1203 |
+
"learning_rate": 1.0155424063052306e-05,
|
| 1204 |
+
"loss": 0.3581,
|
| 1205 |
+
"loss_nan_ranks": 0,
|
| 1206 |
+
"loss_rank_avg": 0.044180020689964294,
|
| 1207 |
+
"step": 545,
|
| 1208 |
+
"valid_targets_mean": 1112.9,
|
| 1209 |
+
"valid_targets_min": 284
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 3.528180354267311,
|
| 1213 |
+
"grad_norm": 0.5865778091291848,
|
| 1214 |
+
"learning_rate": 9.768369827063803e-06,
|
| 1215 |
+
"loss": 0.375,
|
| 1216 |
+
"loss_nan_ranks": 0,
|
| 1217 |
+
"loss_rank_avg": 0.04223884642124176,
|
| 1218 |
+
"step": 550,
|
| 1219 |
+
"valid_targets_mean": 846.5,
|
| 1220 |
+
"valid_targets_min": 231
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.5603864734299515,
|
| 1224 |
+
"grad_norm": 0.484120177701455,
|
| 1225 |
+
"learning_rate": 9.386438212284372e-06,
|
| 1226 |
+
"loss": 0.3564,
|
| 1227 |
+
"loss_nan_ranks": 0,
|
| 1228 |
+
"loss_rank_avg": 0.0398784764111042,
|
| 1229 |
+
"step": 555,
|
| 1230 |
+
"valid_targets_mean": 1083.4,
|
| 1231 |
+
"valid_targets_min": 397
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 3.5925925925925926,
|
| 1235 |
+
"grad_norm": 0.5654014733466529,
|
| 1236 |
+
"learning_rate": 9.009820438583881e-06,
|
| 1237 |
+
"loss": 0.3629,
|
| 1238 |
+
"loss_nan_ranks": 0,
|
| 1239 |
+
"loss_rank_avg": 0.06269536912441254,
|
| 1240 |
+
"step": 560,
|
| 1241 |
+
"valid_targets_mean": 954.4,
|
| 1242 |
+
"valid_targets_min": 297
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.6247987117552336,
|
| 1246 |
+
"grad_norm": 0.42473927669630696,
|
| 1247 |
+
"learning_rate": 8.638705065376887e-06,
|
| 1248 |
+
"loss": 0.3464,
|
| 1249 |
+
"loss_nan_ranks": 0,
|
| 1250 |
+
"loss_rank_avg": 0.021862631663680077,
|
| 1251 |
+
"step": 565,
|
| 1252 |
+
"valid_targets_mean": 506.1,
|
| 1253 |
+
"valid_targets_min": 266
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 3.6570048309178746,
|
| 1257 |
+
"grad_norm": 0.3862733964555012,
|
| 1258 |
+
"learning_rate": 8.27327789721746e-06,
|
| 1259 |
+
"loss": 0.3585,
|
| 1260 |
+
"loss_nan_ranks": 0,
|
| 1261 |
+
"loss_rank_avg": 0.04623720049858093,
|
| 1262 |
+
"step": 570,
|
| 1263 |
+
"valid_targets_mean": 1697.0,
|
| 1264 |
+
"valid_targets_min": 298
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 3.6892109500805152,
|
| 1268 |
+
"grad_norm": 0.4125076115623328,
|
| 1269 |
+
"learning_rate": 7.913721890773354e-06,
|
| 1270 |
+
"loss": 0.3258,
|
| 1271 |
+
"loss_nan_ranks": 0,
|
| 1272 |
+
"loss_rank_avg": 0.0626937747001648,
|
| 1273 |
+
"step": 575,
|
| 1274 |
+
"valid_targets_mean": 2128.5,
|
| 1275 |
+
"valid_targets_min": 338
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.7214170692431563,
|
| 1279 |
+
"grad_norm": 0.5688902989190885,
|
| 1280 |
+
"learning_rate": 7.560217063225923e-06,
|
| 1281 |
+
"loss": 0.3681,
|
| 1282 |
+
"loss_nan_ranks": 0,
|
| 1283 |
+
"loss_rank_avg": 0.05661293864250183,
|
| 1284 |
+
"step": 580,
|
| 1285 |
+
"valid_targets_mean": 855.2,
|
| 1286 |
+
"valid_targets_min": 322
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.753623188405797,
|
| 1290 |
+
"grad_norm": 0.47371336118949064,
|
| 1291 |
+
"learning_rate": 7.212940402141808e-06,
|
| 1292 |
+
"loss": 0.3543,
|
| 1293 |
+
"loss_nan_ranks": 0,
|
| 1294 |
+
"loss_rank_avg": 0.032223887741565704,
|
| 1295 |
+
"step": 585,
|
| 1296 |
+
"valid_targets_mean": 570.1,
|
| 1297 |
+
"valid_targets_min": 331
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.785829307568438,
|
| 1301 |
+
"grad_norm": 0.4522966191465321,
|
| 1302 |
+
"learning_rate": 6.8720657768614115e-06,
|
| 1303 |
+
"loss": 0.3299,
|
| 1304 |
+
"loss_nan_ranks": 0,
|
| 1305 |
+
"loss_rank_avg": 0.035974081605672836,
|
| 1306 |
+
"step": 590,
|
| 1307 |
+
"valid_targets_mean": 813.8,
|
| 1308 |
+
"valid_targets_min": 331
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.818035426731079,
|
| 1312 |
+
"grad_norm": 0.4652139120483709,
|
| 1313 |
+
"learning_rate": 6.537763851448593e-06,
|
| 1314 |
+
"loss": 0.3628,
|
| 1315 |
+
"loss_nan_ranks": 0,
|
| 1316 |
+
"loss_rank_avg": 0.04356799274682999,
|
| 1317 |
+
"step": 595,
|
| 1318 |
+
"valid_targets_mean": 1215.0,
|
| 1319 |
+
"valid_targets_min": 288
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.85024154589372,
|
| 1323 |
+
"grad_norm": 0.46874463875570954,
|
| 1324 |
+
"learning_rate": 6.210201999245087e-06,
|
| 1325 |
+
"loss": 0.3436,
|
| 1326 |
+
"loss_nan_ranks": 0,
|
| 1327 |
+
"loss_rank_avg": 0.029137909412384033,
|
| 1328 |
+
"step": 600,
|
| 1329 |
+
"valid_targets_mean": 621.4,
|
| 1330 |
+
"valid_targets_min": 278
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 3.8824476650563606,
|
| 1334 |
+
"grad_norm": 0.41711178274476524,
|
| 1335 |
+
"learning_rate": 5.889544219072465e-06,
|
| 1336 |
+
"loss": 0.3406,
|
| 1337 |
+
"loss_nan_ranks": 0,
|
| 1338 |
+
"loss_rank_avg": 0.045179564505815506,
|
| 1339 |
+
"step": 605,
|
| 1340 |
+
"valid_targets_mean": 1414.2,
|
| 1341 |
+
"valid_targets_min": 342
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 3.9146537842190017,
|
| 1345 |
+
"grad_norm": 0.36858876028850934,
|
| 1346 |
+
"learning_rate": 5.5759510531237116e-06,
|
| 1347 |
+
"loss": 0.3473,
|
| 1348 |
+
"loss_nan_ranks": 0,
|
| 1349 |
+
"loss_rank_avg": 0.03258288651704788,
|
| 1350 |
+
"step": 610,
|
| 1351 |
+
"valid_targets_mean": 1184.9,
|
| 1352 |
+
"valid_targets_min": 350
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.9468599033816423,
|
| 1356 |
+
"grad_norm": 0.3655458437589055,
|
| 1357 |
+
"learning_rate": 5.269579506585259e-06,
|
| 1358 |
+
"loss": 0.3383,
|
| 1359 |
+
"loss_nan_ranks": 0,
|
| 1360 |
+
"loss_rank_avg": 0.024550924077630043,
|
| 1361 |
+
"step": 615,
|
| 1362 |
+
"valid_targets_mean": 761.0,
|
| 1363 |
+
"valid_targets_min": 367
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 3.9790660225442833,
|
| 1367 |
+
"grad_norm": 0.5867554039598152,
|
| 1368 |
+
"learning_rate": 4.97058296903e-06,
|
| 1369 |
+
"loss": 0.3549,
|
| 1370 |
+
"loss_nan_ranks": 0,
|
| 1371 |
+
"loss_rank_avg": 0.04499742016196251,
|
| 1372 |
+
"step": 620,
|
| 1373 |
+
"valid_targets_mean": 819.9,
|
| 1374 |
+
"valid_targets_min": 312
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.006441223832528,
|
| 1378 |
+
"grad_norm": 0.5337919310243177,
|
| 1379 |
+
"learning_rate": 4.679111137620442e-06,
|
| 1380 |
+
"loss": 0.3519,
|
| 1381 |
+
"loss_nan_ranks": 0,
|
| 1382 |
+
"loss_rank_avg": 0.04486808925867081,
|
| 1383 |
+
"step": 625,
|
| 1384 |
+
"valid_targets_mean": 827.9,
|
| 1385 |
+
"valid_targets_min": 261
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.038647342995169,
|
| 1389 |
+
"grad_norm": 0.5000808046641099,
|
| 1390 |
+
"learning_rate": 4.395309942160551e-06,
|
| 1391 |
+
"loss": 0.3509,
|
| 1392 |
+
"loss_nan_ranks": 0,
|
| 1393 |
+
"loss_rank_avg": 0.03933977708220482,
|
| 1394 |
+
"step": 630,
|
| 1395 |
+
"valid_targets_mean": 830.9,
|
| 1396 |
+
"valid_targets_min": 299
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.07085346215781,
|
| 1400 |
+
"grad_norm": 0.4557917329913038,
|
| 1401 |
+
"learning_rate": 4.119321472033779e-06,
|
| 1402 |
+
"loss": 0.3471,
|
| 1403 |
+
"loss_nan_ranks": 0,
|
| 1404 |
+
"loss_rank_avg": 0.054948341101408005,
|
| 1405 |
+
"step": 635,
|
| 1406 |
+
"valid_targets_mean": 1991.6,
|
| 1407 |
+
"valid_targets_min": 267
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.1030595813204505,
|
| 1411 |
+
"grad_norm": 0.4721247455404434,
|
| 1412 |
+
"learning_rate": 3.851283905063863e-06,
|
| 1413 |
+
"loss": 0.3463,
|
| 1414 |
+
"loss_nan_ranks": 0,
|
| 1415 |
+
"loss_rank_avg": 0.04883628711104393,
|
| 1416 |
+
"step": 640,
|
| 1417 |
+
"valid_targets_mean": 1025.4,
|
| 1418 |
+
"valid_targets_min": 344
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.1352657004830915,
|
| 1422 |
+
"grad_norm": 0.44811926966002247,
|
| 1423 |
+
"learning_rate": 3.5913314383339937e-06,
|
| 1424 |
+
"loss": 0.3443,
|
| 1425 |
+
"loss_nan_ranks": 0,
|
| 1426 |
+
"loss_rank_avg": 0.02158988080918789,
|
| 1427 |
+
"step": 645,
|
| 1428 |
+
"valid_targets_mean": 505.0,
|
| 1429 |
+
"valid_targets_min": 200
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 4.1674718196457325,
|
| 1433 |
+
"grad_norm": 0.4830380662288485,
|
| 1434 |
+
"learning_rate": 3.3395942209989694e-06,
|
| 1435 |
+
"loss": 0.3395,
|
| 1436 |
+
"loss_nan_ranks": 0,
|
| 1437 |
+
"loss_rank_avg": 0.03771327808499336,
|
| 1438 |
+
"step": 650,
|
| 1439 |
+
"valid_targets_mean": 780.9,
|
| 1440 |
+
"valid_targets_min": 342
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 4.199677938808374,
|
| 1444 |
+
"grad_norm": 0.4975062818136211,
|
| 1445 |
+
"learning_rate": 3.0961982891241083e-06,
|
| 1446 |
+
"loss": 0.338,
|
| 1447 |
+
"loss_nan_ranks": 0,
|
| 1448 |
+
"loss_rank_avg": 0.04027272388339043,
|
| 1449 |
+
"step": 655,
|
| 1450 |
+
"valid_targets_mean": 965.4,
|
| 1451 |
+
"valid_targets_min": 331
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.231884057971015,
|
| 1455 |
+
"grad_norm": 0.5186580520002718,
|
| 1456 |
+
"learning_rate": 2.8612655025833148e-06,
|
| 1457 |
+
"loss": 0.3311,
|
| 1458 |
+
"loss_nan_ranks": 0,
|
| 1459 |
+
"loss_rank_avg": 0.0684180036187172,
|
| 1460 |
+
"step": 660,
|
| 1461 |
+
"valid_targets_mean": 1254.0,
|
| 1462 |
+
"valid_targets_min": 253
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.264090177133656,
|
| 1466 |
+
"grad_norm": 0.5925008943137404,
|
| 1467 |
+
"learning_rate": 2.6349134840481294e-06,
|
| 1468 |
+
"loss": 0.3621,
|
| 1469 |
+
"loss_nan_ranks": 0,
|
| 1470 |
+
"loss_rank_avg": 0.05215730518102646,
|
| 1471 |
+
"step": 665,
|
| 1472 |
+
"valid_targets_mean": 731.0,
|
| 1473 |
+
"valid_targets_min": 272
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.296296296296296,
|
| 1477 |
+
"grad_norm": 0.41226419968489375,
|
| 1478 |
+
"learning_rate": 2.417255560098111e-06,
|
| 1479 |
+
"loss": 0.341,
|
| 1480 |
+
"loss_nan_ranks": 0,
|
| 1481 |
+
"loss_rank_avg": 0.03634045645594597,
|
| 1482 |
+
"step": 670,
|
| 1483 |
+
"valid_targets_mean": 1005.0,
|
| 1484 |
+
"valid_targets_min": 357
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.328502415458937,
|
| 1488 |
+
"grad_norm": 0.4129907136798245,
|
| 1489 |
+
"learning_rate": 2.2084007044821764e-06,
|
| 1490 |
+
"loss": 0.3291,
|
| 1491 |
+
"loss_nan_ranks": 0,
|
| 1492 |
+
"loss_rank_avg": 0.03767111897468567,
|
| 1493 |
+
"step": 675,
|
| 1494 |
+
"valid_targets_mean": 1171.0,
|
| 1495 |
+
"valid_targets_min": 289
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.360708534621578,
|
| 1499 |
+
"grad_norm": 0.5348812254442832,
|
| 1500 |
+
"learning_rate": 2.0084534835592584e-06,
|
| 1501 |
+
"loss": 0.3387,
|
| 1502 |
+
"loss_nan_ranks": 0,
|
| 1503 |
+
"loss_rank_avg": 0.04408322647213936,
|
| 1504 |
+
"step": 680,
|
| 1505 |
+
"valid_targets_mean": 861.4,
|
| 1506 |
+
"valid_targets_min": 346
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.392914653784219,
|
| 1510 |
+
"grad_norm": 0.4983477080423943,
|
| 1511 |
+
"learning_rate": 1.817514003945524e-06,
|
| 1512 |
+
"loss": 0.3446,
|
| 1513 |
+
"loss_nan_ranks": 0,
|
| 1514 |
+
"loss_rank_avg": 0.03348889201879501,
|
| 1515 |
+
"step": 685,
|
| 1516 |
+
"valid_targets_mean": 840.8,
|
| 1517 |
+
"valid_targets_min": 301
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 4.42512077294686,
|
| 1521 |
+
"grad_norm": 0.5216601301709456,
|
| 1522 |
+
"learning_rate": 1.6356778623945223e-06,
|
| 1523 |
+
"loss": 0.3368,
|
| 1524 |
+
"loss_nan_ranks": 0,
|
| 1525 |
+
"loss_rank_avg": 0.034096863120794296,
|
| 1526 |
+
"step": 690,
|
| 1527 |
+
"valid_targets_mean": 715.5,
|
| 1528 |
+
"valid_targets_min": 244
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.457326892109501,
|
| 1532 |
+
"grad_norm": 0.4655020726002043,
|
| 1533 |
+
"learning_rate": 1.4630360979351644e-06,
|
| 1534 |
+
"loss": 0.3668,
|
| 1535 |
+
"loss_nan_ranks": 0,
|
| 1536 |
+
"loss_rank_avg": 0.07300788164138794,
|
| 1537 |
+
"step": 695,
|
| 1538 |
+
"valid_targets_mean": 2403.8,
|
| 1539 |
+
"valid_targets_min": 313
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.489533011272142,
|
| 1543 |
+
"grad_norm": 0.5310053686779203,
|
| 1544 |
+
"learning_rate": 1.2996751462917057e-06,
|
| 1545 |
+
"loss": 0.3544,
|
| 1546 |
+
"loss_nan_ranks": 0,
|
| 1547 |
+
"loss_rank_avg": 0.037691399455070496,
|
| 1548 |
+
"step": 700,
|
| 1549 |
+
"valid_targets_mean": 850.4,
|
| 1550 |
+
"valid_targets_min": 306
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.521739130434782,
|
| 1554 |
+
"grad_norm": 0.4310213806844335,
|
| 1555 |
+
"learning_rate": 1.1456767966083393e-06,
|
| 1556 |
+
"loss": 0.3441,
|
| 1557 |
+
"loss_nan_ranks": 0,
|
| 1558 |
+
"loss_rank_avg": 0.03169054538011551,
|
| 1559 |
+
"step": 705,
|
| 1560 |
+
"valid_targets_mean": 730.5,
|
| 1561 |
+
"valid_targets_min": 316
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 4.553945249597423,
|
| 1565 |
+
"grad_norm": 0.40942357788440725,
|
| 1566 |
+
"learning_rate": 1.0011181505002488e-06,
|
| 1567 |
+
"loss": 0.3616,
|
| 1568 |
+
"loss_nan_ranks": 0,
|
| 1569 |
+
"loss_rank_avg": 0.02835988998413086,
|
| 1570 |
+
"step": 710,
|
| 1571 |
+
"valid_targets_mean": 966.9,
|
| 1572 |
+
"valid_targets_min": 248
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.586151368760064,
|
| 1576 |
+
"grad_norm": 0.4064126072802411,
|
| 1577 |
+
"learning_rate": 8.660715834514977e-07,
|
| 1578 |
+
"loss": 0.3394,
|
| 1579 |
+
"loss_nan_ranks": 0,
|
| 1580 |
+
"loss_rank_avg": 0.019462428987026215,
|
| 1581 |
+
"step": 715,
|
| 1582 |
+
"valid_targets_mean": 683.4,
|
| 1583 |
+
"valid_targets_min": 303
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 4.618357487922705,
|
| 1587 |
+
"grad_norm": 0.4659035655355515,
|
| 1588 |
+
"learning_rate": 7.406047085791379e-07,
|
| 1589 |
+
"loss": 0.3367,
|
| 1590 |
+
"loss_nan_ranks": 0,
|
| 1591 |
+
"loss_rank_avg": 0.04081908240914345,
|
| 1592 |
+
"step": 720,
|
| 1593 |
+
"valid_targets_mean": 1303.8,
|
| 1594 |
+
"valid_targets_min": 284
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 4.650563607085346,
|
| 1598 |
+
"grad_norm": 0.5553298528236867,
|
| 1599 |
+
"learning_rate": 6.247803427816945e-07,
|
| 1600 |
+
"loss": 0.3613,
|
| 1601 |
+
"loss_nan_ranks": 0,
|
| 1602 |
+
"loss_rank_avg": 0.0552799291908741,
|
| 1603 |
+
"step": 725,
|
| 1604 |
+
"valid_targets_mean": 973.4,
|
| 1605 |
+
"valid_targets_min": 357
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.6827697262479875,
|
| 1609 |
+
"grad_norm": 0.5045394202661283,
|
| 1610 |
+
"learning_rate": 5.186564752889078e-07,
|
| 1611 |
+
"loss": 0.3547,
|
| 1612 |
+
"loss_nan_ranks": 0,
|
| 1613 |
+
"loss_rank_avg": 0.08811699599027634,
|
| 1614 |
+
"step": 730,
|
| 1615 |
+
"valid_targets_mean": 1683.6,
|
| 1616 |
+
"valid_targets_min": 439
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.714975845410628,
|
| 1620 |
+
"grad_norm": 0.41108260312594874,
|
| 1621 |
+
"learning_rate": 4.22286238628562e-07,
|
| 1622 |
+
"loss": 0.3564,
|
| 1623 |
+
"loss_nan_ranks": 0,
|
| 1624 |
+
"loss_rank_avg": 0.04261046648025513,
|
| 1625 |
+
"step": 735,
|
| 1626 |
+
"valid_targets_mean": 1162.9,
|
| 1627 |
+
"valid_targets_min": 225
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.747181964573269,
|
| 1631 |
+
"grad_norm": 0.37863905329842407,
|
| 1632 |
+
"learning_rate": 3.357178820248641e-07,
|
| 1633 |
+
"loss": 0.3402,
|
| 1634 |
+
"loss_nan_ranks": 0,
|
| 1635 |
+
"loss_rank_avg": 0.06985923647880554,
|
| 1636 |
+
"step": 740,
|
| 1637 |
+
"valid_targets_mean": 3784.5,
|
| 1638 |
+
"valid_targets_min": 348
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 4.77938808373591,
|
| 1642 |
+
"grad_norm": 0.38057216907053537,
|
| 1643 |
+
"learning_rate": 2.5899474724174313e-07,
|
| 1644 |
+
"loss": 0.3272,
|
| 1645 |
+
"loss_nan_ranks": 0,
|
| 1646 |
+
"loss_rank_avg": 0.024681851267814636,
|
| 1647 |
+
"step": 745,
|
| 1648 |
+
"valid_targets_mean": 856.8,
|
| 1649 |
+
"valid_targets_min": 320
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.811594202898551,
|
| 1653 |
+
"grad_norm": 0.4980702323011905,
|
| 1654 |
+
"learning_rate": 1.9215524688313893e-07,
|
| 1655 |
+
"loss": 0.3422,
|
| 1656 |
+
"loss_nan_ranks": 0,
|
| 1657 |
+
"loss_rank_avg": 0.03393610566854477,
|
| 1658 |
+
"step": 750,
|
| 1659 |
+
"valid_targets_mean": 697.4,
|
| 1660 |
+
"valid_targets_min": 304
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 4.843800322061192,
|
| 1664 |
+
"grad_norm": 0.48534143807405106,
|
| 1665 |
+
"learning_rate": 1.3523284516113955e-07,
|
| 1666 |
+
"loss": 0.3442,
|
| 1667 |
+
"loss_nan_ranks": 0,
|
| 1668 |
+
"loss_rank_avg": 0.04657585173845291,
|
| 1669 |
+
"step": 755,
|
| 1670 |
+
"valid_targets_mean": 1234.6,
|
| 1671 |
+
"valid_targets_min": 294
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 4.876006441223833,
|
| 1675 |
+
"grad_norm": 0.4327518473258517,
|
| 1676 |
+
"learning_rate": 8.825604114162556e-08,
|
| 1677 |
+
"loss": 0.3246,
|
| 1678 |
+
"loss_nan_ranks": 0,
|
| 1679 |
+
"loss_rank_avg": 0.03470573574304581,
|
| 1680 |
+
"step": 760,
|
| 1681 |
+
"valid_targets_mean": 690.1,
|
| 1682 |
+
"valid_targets_min": 276
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.908212560386474,
|
| 1686 |
+
"grad_norm": 0.5175651819741588,
|
| 1687 |
+
"learning_rate": 5.1248354475768034e-08,
|
| 1688 |
+
"loss": 0.3555,
|
| 1689 |
+
"loss_nan_ranks": 0,
|
| 1690 |
+
"loss_rank_avg": 0.05181993171572685,
|
| 1691 |
+
"step": 765,
|
| 1692 |
+
"valid_targets_mean": 1085.9,
|
| 1693 |
+
"valid_targets_min": 303
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 4.940418679549114,
|
| 1697 |
+
"grad_norm": 0.5505526615345873,
|
| 1698 |
+
"learning_rate": 2.4228313624568013e-08,
|
| 1699 |
+
"loss": 0.3575,
|
| 1700 |
+
"loss_nan_ranks": 0,
|
| 1701 |
+
"loss_rank_avg": 0.03321858495473862,
|
| 1702 |
+
"step": 770,
|
| 1703 |
+
"valid_targets_mean": 728.8,
|
| 1704 |
+
"valid_targets_min": 168
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.972624798711755,
|
| 1708 |
+
"grad_norm": 0.5754568951885141,
|
| 1709 |
+
"learning_rate": 7.209446582292501e-09,
|
| 1710 |
+
"loss": 0.3547,
|
| 1711 |
+
"loss_nan_ranks": 0,
|
| 1712 |
+
"loss_rank_avg": 0.06288649886846542,
|
| 1713 |
+
"step": 775,
|
| 1714 |
+
"valid_targets_mean": 1244.1,
|
| 1715 |
+
"valid_targets_min": 286
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 5.0,
|
| 1719 |
+
"grad_norm": 1.1816753894303154,
|
| 1720 |
+
"learning_rate": 2.002741034812239e-10,
|
| 1721 |
+
"loss": 0.3613,
|
| 1722 |
+
"loss_nan_ranks": 0,
|
| 1723 |
+
"loss_rank_avg": 0.23872976005077362,
|
| 1724 |
+
"step": 780,
|
| 1725 |
+
"valid_targets_mean": 1017.1,
|
| 1726 |
+
"valid_targets_min": 327
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.0,
|
| 1730 |
+
"loss_nan_ranks": 0,
|
| 1731 |
+
"loss_rank_avg": 0.23872976005077362,
|
| 1732 |
+
"step": 780,
|
| 1733 |
+
"total_flos": 4.410167938263286e+17,
|
| 1734 |
+
"train_loss": 0.3955344903163421,
|
| 1735 |
+
"train_runtime": 13041.82,
|
| 1736 |
+
"train_samples_per_second": 3.807,
|
| 1737 |
+
"train_steps_per_second": 0.06,
|
| 1738 |
+
"valid_targets_mean": 1017.1,
|
| 1739 |
+
"valid_targets_min": 327
|
| 1740 |
+
}
|
| 1741 |
+
],
|
| 1742 |
+
"logging_steps": 5,
|
| 1743 |
+
"max_steps": 780,
|
| 1744 |
+
"num_input_tokens_seen": 0,
|
| 1745 |
+
"num_train_epochs": 5,
|
| 1746 |
+
"save_steps": 500,
|
| 1747 |
+
"stateful_callbacks": {
|
| 1748 |
+
"TrainerControl": {
|
| 1749 |
+
"args": {
|
| 1750 |
+
"should_epoch_stop": false,
|
| 1751 |
+
"should_evaluate": false,
|
| 1752 |
+
"should_log": false,
|
| 1753 |
+
"should_save": false,
|
| 1754 |
+
"should_training_stop": false
|
| 1755 |
+
},
|
| 1756 |
+
"attributes": {}
|
| 1757 |
+
}
|
| 1758 |
+
},
|
| 1759 |
+
"total_flos": 4.410167938263286e+17,
|
| 1760 |
+
"train_batch_size": 1,
|
| 1761 |
+
"trial_name": null,
|
| 1762 |
+
"trial_params": null
|
| 1763 |
+
}
|
training_loss.png
ADDED
|