Model save
Browse files- README.md +67 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- train_results.json +9 -0
- trainer_state.json +2651 -0
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: allenai/OLMo-1B-hf
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- generator
|
| 10 |
+
model-index:
|
| 11 |
+
- name: sft-olmo-1b
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
+
|
| 18 |
+
# sft-olmo-1b
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [allenai/OLMo-1B-hf](https://huggingface.co/allenai/OLMo-1B-hf) on the generator dataset.
|
| 21 |
+
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 0.8224
|
| 23 |
+
|
| 24 |
+
## Model description
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Intended uses & limitations
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training and evaluation data
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training procedure
|
| 37 |
+
|
| 38 |
+
### Training hyperparameters
|
| 39 |
+
|
| 40 |
+
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 2e-05
|
| 42 |
+
- train_batch_size: 8
|
| 43 |
+
- eval_batch_size: 2
|
| 44 |
+
- seed: 42
|
| 45 |
+
- distributed_type: multi-GPU
|
| 46 |
+
- gradient_accumulation_steps: 16
|
| 47 |
+
- total_train_batch_size: 128
|
| 48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
+
- lr_scheduler_type: cosine
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.03
|
| 51 |
+
- num_epochs: 3
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 56 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 57 |
+
| 1.1754 | 0.9992 | 1236 | 1.0556 |
|
| 58 |
+
| 0.9628 | 1.9993 | 2473 | 0.8751 |
|
| 59 |
+
| 0.801 | 2.9977 | 3708 | 0.8224 |
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
### Framework versions
|
| 63 |
+
|
| 64 |
+
- Transformers 4.40.0
|
| 65 |
+
- Pytorch 2.1.2
|
| 66 |
+
- Datasets 2.14.6
|
| 67 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.997726239199636,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 1.0273753281164324,
|
| 5 |
+
"train_runtime": 58675.1239,
|
| 6 |
+
"train_samples": 326149,
|
| 7 |
+
"train_samples_per_second": 8.095,
|
| 8 |
+
"train_steps_per_second": 0.063
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"eos_token_id": 0,
|
| 4 |
+
"pad_token_id": 1,
|
| 5 |
+
"transformers_version": "4.40.0"
|
| 6 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.997726239199636,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 1.0273753281164324,
|
| 5 |
+
"train_runtime": 58675.1239,
|
| 6 |
+
"train_samples": 326149,
|
| 7 |
+
"train_samples_per_second": 8.095,
|
| 8 |
+
"train_steps_per_second": 0.063
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2651 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.997726239199636,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3708,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0008084482845737962,
|
| 13 |
+
"grad_norm": 10.217898134986712,
|
| 14 |
+
"learning_rate": 1.7857142857142858e-07,
|
| 15 |
+
"loss": 1.7929,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.008084482845737961,
|
| 20 |
+
"grad_norm": 8.447331849071613,
|
| 21 |
+
"learning_rate": 1.7857142857142859e-06,
|
| 22 |
+
"loss": 1.8062,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016168965691475922,
|
| 27 |
+
"grad_norm": 3.3131831113560053,
|
| 28 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 29 |
+
"loss": 1.6026,
|
| 30 |
+
"step": 20
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.024253448537213885,
|
| 34 |
+
"grad_norm": 1.9123288015392683,
|
| 35 |
+
"learning_rate": 5.357142857142857e-06,
|
| 36 |
+
"loss": 1.5044,
|
| 37 |
+
"step": 30
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.032337931382951844,
|
| 41 |
+
"grad_norm": 1.5061470360359468,
|
| 42 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 43 |
+
"loss": 1.46,
|
| 44 |
+
"step": 40
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04042241422868981,
|
| 48 |
+
"grad_norm": 1.3730571252951111,
|
| 49 |
+
"learning_rate": 8.92857142857143e-06,
|
| 50 |
+
"loss": 1.4242,
|
| 51 |
+
"step": 50
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04850689707442777,
|
| 55 |
+
"grad_norm": 1.3696793978053872,
|
| 56 |
+
"learning_rate": 1.0714285714285714e-05,
|
| 57 |
+
"loss": 1.3962,
|
| 58 |
+
"step": 60
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05659137992016573,
|
| 62 |
+
"grad_norm": 1.3254100387294059,
|
| 63 |
+
"learning_rate": 1.25e-05,
|
| 64 |
+
"loss": 1.4019,
|
| 65 |
+
"step": 70
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06467586276590369,
|
| 69 |
+
"grad_norm": 1.300617283886077,
|
| 70 |
+
"learning_rate": 1.4285714285714287e-05,
|
| 71 |
+
"loss": 1.38,
|
| 72 |
+
"step": 80
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07276034561164166,
|
| 76 |
+
"grad_norm": 1.2890020177179862,
|
| 77 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 78 |
+
"loss": 1.3692,
|
| 79 |
+
"step": 90
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08084482845737961,
|
| 83 |
+
"grad_norm": 1.3617464980600442,
|
| 84 |
+
"learning_rate": 1.785714285714286e-05,
|
| 85 |
+
"loss": 1.3717,
|
| 86 |
+
"step": 100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08892931130311758,
|
| 90 |
+
"grad_norm": 1.3364658238171299,
|
| 91 |
+
"learning_rate": 1.9642857142857145e-05,
|
| 92 |
+
"loss": 1.3445,
|
| 93 |
+
"step": 110
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09701379414885554,
|
| 97 |
+
"grad_norm": 1.406654933905743,
|
| 98 |
+
"learning_rate": 1.999975576461237e-05,
|
| 99 |
+
"loss": 1.3724,
|
| 100 |
+
"step": 120
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.1050982769945935,
|
| 104 |
+
"grad_norm": 1.4611284363694028,
|
| 105 |
+
"learning_rate": 1.999876357879684e-05,
|
| 106 |
+
"loss": 1.3446,
|
| 107 |
+
"step": 130
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.11318275984033146,
|
| 111 |
+
"grad_norm": 1.3711888479566774,
|
| 112 |
+
"learning_rate": 1.9997008253510416e-05,
|
| 113 |
+
"loss": 1.3515,
|
| 114 |
+
"step": 140
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.12126724268606942,
|
| 118 |
+
"grad_norm": 1.409811295994504,
|
| 119 |
+
"learning_rate": 1.9994489922725454e-05,
|
| 120 |
+
"loss": 1.342,
|
| 121 |
+
"step": 150
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.12935172553180738,
|
| 125 |
+
"grad_norm": 1.3918093353079757,
|
| 126 |
+
"learning_rate": 1.9991208778649485e-05,
|
| 127 |
+
"loss": 1.3493,
|
| 128 |
+
"step": 160
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.13743620837754536,
|
| 132 |
+
"grad_norm": 1.3597096387911873,
|
| 133 |
+
"learning_rate": 1.998716507171053e-05,
|
| 134 |
+
"loss": 1.3186,
|
| 135 |
+
"step": 170
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.14552069122328332,
|
| 139 |
+
"grad_norm": 1.3532852407233782,
|
| 140 |
+
"learning_rate": 1.998235911053798e-05,
|
| 141 |
+
"loss": 1.3426,
|
| 142 |
+
"step": 180
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.15360517406902127,
|
| 146 |
+
"grad_norm": 1.395929355438092,
|
| 147 |
+
"learning_rate": 1.9976791261939064e-05,
|
| 148 |
+
"loss": 1.338,
|
| 149 |
+
"step": 190
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.16168965691475923,
|
| 153 |
+
"grad_norm": 1.3346897807354206,
|
| 154 |
+
"learning_rate": 1.997046195087082e-05,
|
| 155 |
+
"loss": 1.3209,
|
| 156 |
+
"step": 200
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.16977413976049718,
|
| 160 |
+
"grad_norm": 1.3501622187451188,
|
| 161 |
+
"learning_rate": 1.996337166040769e-05,
|
| 162 |
+
"loss": 1.3279,
|
| 163 |
+
"step": 210
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.17785862260623517,
|
| 167 |
+
"grad_norm": 1.262079121683393,
|
| 168 |
+
"learning_rate": 1.995552093170463e-05,
|
| 169 |
+
"loss": 1.3135,
|
| 170 |
+
"step": 220
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.18594310545197312,
|
| 174 |
+
"grad_norm": 1.324883144834464,
|
| 175 |
+
"learning_rate": 1.994691036395583e-05,
|
| 176 |
+
"loss": 1.306,
|
| 177 |
+
"step": 230
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.19402758829771108,
|
| 181 |
+
"grad_norm": 1.373867783740766,
|
| 182 |
+
"learning_rate": 1.9937540614348944e-05,
|
| 183 |
+
"loss": 1.3018,
|
| 184 |
+
"step": 240
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.20211207114344903,
|
| 188 |
+
"grad_norm": 1.4020861161362783,
|
| 189 |
+
"learning_rate": 1.992741239801498e-05,
|
| 190 |
+
"loss": 1.3203,
|
| 191 |
+
"step": 250
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.210196553989187,
|
| 195 |
+
"grad_norm": 1.3484650757297245,
|
| 196 |
+
"learning_rate": 1.9916526487973678e-05,
|
| 197 |
+
"loss": 1.2939,
|
| 198 |
+
"step": 260
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.21828103683492497,
|
| 202 |
+
"grad_norm": 1.3330965331306333,
|
| 203 |
+
"learning_rate": 1.9904883715074525e-05,
|
| 204 |
+
"loss": 1.2795,
|
| 205 |
+
"step": 270
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.22636551968066293,
|
| 209 |
+
"grad_norm": 1.3917589397233552,
|
| 210 |
+
"learning_rate": 1.989248496793335e-05,
|
| 211 |
+
"loss": 1.269,
|
| 212 |
+
"step": 280
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.23445000252640089,
|
| 216 |
+
"grad_norm": 1.3905412148367542,
|
| 217 |
+
"learning_rate": 1.9879331192864492e-05,
|
| 218 |
+
"loss": 1.286,
|
| 219 |
+
"step": 290
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.24253448537213884,
|
| 223 |
+
"grad_norm": 1.4569325197708967,
|
| 224 |
+
"learning_rate": 1.9865423393808573e-05,
|
| 225 |
+
"loss": 1.2944,
|
| 226 |
+
"step": 300
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2506189682178768,
|
| 230 |
+
"grad_norm": 1.3399495909594208,
|
| 231 |
+
"learning_rate": 1.985076263225588e-05,
|
| 232 |
+
"loss": 1.3106,
|
| 233 |
+
"step": 310
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.25870345106361475,
|
| 237 |
+
"grad_norm": 1.478802579336813,
|
| 238 |
+
"learning_rate": 1.9835350027165342e-05,
|
| 239 |
+
"loss": 1.2994,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.26678793390935274,
|
| 244 |
+
"grad_norm": 1.3105244819439577,
|
| 245 |
+
"learning_rate": 1.9819186754879137e-05,
|
| 246 |
+
"loss": 1.2871,
|
| 247 |
+
"step": 330
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2748724167550907,
|
| 251 |
+
"grad_norm": 1.3667119896120177,
|
| 252 |
+
"learning_rate": 1.9802274049032898e-05,
|
| 253 |
+
"loss": 1.2893,
|
| 254 |
+
"step": 340
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.28295689960082865,
|
| 258 |
+
"grad_norm": 1.5054526064910085,
|
| 259 |
+
"learning_rate": 1.9784613200461568e-05,
|
| 260 |
+
"loss": 1.2912,
|
| 261 |
+
"step": 350
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.29104138244656663,
|
| 265 |
+
"grad_norm": 1.3163243486049039,
|
| 266 |
+
"learning_rate": 1.976620555710087e-05,
|
| 267 |
+
"loss": 1.2761,
|
| 268 |
+
"step": 360
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.29912586529230456,
|
| 272 |
+
"grad_norm": 1.322920539242633,
|
| 273 |
+
"learning_rate": 1.9747052523884435e-05,
|
| 274 |
+
"loss": 1.2572,
|
| 275 |
+
"step": 370
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.30721034813804254,
|
| 279 |
+
"grad_norm": 1.3954468357326724,
|
| 280 |
+
"learning_rate": 1.972715556263657e-05,
|
| 281 |
+
"loss": 1.2745,
|
| 282 |
+
"step": 380
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.3152948309837805,
|
| 286 |
+
"grad_norm": 1.3451929159695755,
|
| 287 |
+
"learning_rate": 1.9706516191960687e-05,
|
| 288 |
+
"loss": 1.2472,
|
| 289 |
+
"step": 390
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.32337931382951846,
|
| 293 |
+
"grad_norm": 1.2765565775996142,
|
| 294 |
+
"learning_rate": 1.9685135987123396e-05,
|
| 295 |
+
"loss": 1.255,
|
| 296 |
+
"step": 400
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.33146379667525644,
|
| 300 |
+
"grad_norm": 1.4632541877317655,
|
| 301 |
+
"learning_rate": 1.966301657993428e-05,
|
| 302 |
+
"loss": 1.2565,
|
| 303 |
+
"step": 410
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.33954827952099437,
|
| 307 |
+
"grad_norm": 1.3554436136314076,
|
| 308 |
+
"learning_rate": 1.9640159658621344e-05,
|
| 309 |
+
"loss": 1.2593,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.34763276236673235,
|
| 314 |
+
"grad_norm": 1.3154961346767526,
|
| 315 |
+
"learning_rate": 1.9616566967702164e-05,
|
| 316 |
+
"loss": 1.2604,
|
| 317 |
+
"step": 430
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.35571724521247033,
|
| 321 |
+
"grad_norm": 1.3833700211512812,
|
| 322 |
+
"learning_rate": 1.9592240307850748e-05,
|
| 323 |
+
"loss": 1.2625,
|
| 324 |
+
"step": 440
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.36380172805820826,
|
| 328 |
+
"grad_norm": 1.2812641775550833,
|
| 329 |
+
"learning_rate": 1.95671815357601e-05,
|
| 330 |
+
"loss": 1.2661,
|
| 331 |
+
"step": 450
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.37188621090394625,
|
| 335 |
+
"grad_norm": 1.3509908047727408,
|
| 336 |
+
"learning_rate": 1.954139256400049e-05,
|
| 337 |
+
"loss": 1.2448,
|
| 338 |
+
"step": 460
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3799706937496842,
|
| 342 |
+
"grad_norm": 1.356891388574271,
|
| 343 |
+
"learning_rate": 1.951487536087352e-05,
|
| 344 |
+
"loss": 1.2551,
|
| 345 |
+
"step": 470
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.38805517659542216,
|
| 349 |
+
"grad_norm": 1.2921423460134738,
|
| 350 |
+
"learning_rate": 1.948763195026186e-05,
|
| 351 |
+
"loss": 1.2503,
|
| 352 |
+
"step": 480
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.39613965944116014,
|
| 356 |
+
"grad_norm": 1.3494188641362486,
|
| 357 |
+
"learning_rate": 1.9459664411474793e-05,
|
| 358 |
+
"loss": 1.2509,
|
| 359 |
+
"step": 490
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.40422414228689807,
|
| 363 |
+
"grad_norm": 1.336605272931222,
|
| 364 |
+
"learning_rate": 1.9430974879089522e-05,
|
| 365 |
+
"loss": 1.251,
|
| 366 |
+
"step": 500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.41230862513263605,
|
| 370 |
+
"grad_norm": 1.3167568815144604,
|
| 371 |
+
"learning_rate": 1.9401565542788238e-05,
|
| 372 |
+
"loss": 1.2341,
|
| 373 |
+
"step": 510
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.420393107978374,
|
| 377 |
+
"grad_norm": 1.3704112316871029,
|
| 378 |
+
"learning_rate": 1.9371438647191007e-05,
|
| 379 |
+
"loss": 1.2483,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.42847759082411196,
|
| 384 |
+
"grad_norm": 1.2971253486447214,
|
| 385 |
+
"learning_rate": 1.9340596491684443e-05,
|
| 386 |
+
"loss": 1.2483,
|
| 387 |
+
"step": 530
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.43656207366984995,
|
| 391 |
+
"grad_norm": 1.278671915851734,
|
| 392 |
+
"learning_rate": 1.9309041430246228e-05,
|
| 393 |
+
"loss": 1.247,
|
| 394 |
+
"step": 540
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4446465565155879,
|
| 398 |
+
"grad_norm": 1.7143062654632688,
|
| 399 |
+
"learning_rate": 1.927677587126542e-05,
|
| 400 |
+
"loss": 1.2582,
|
| 401 |
+
"step": 550
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.45273103936132586,
|
| 405 |
+
"grad_norm": 3.071400396021207,
|
| 406 |
+
"learning_rate": 1.924380227735867e-05,
|
| 407 |
+
"loss": 1.2369,
|
| 408 |
+
"step": 560
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.46081552220706384,
|
| 412 |
+
"grad_norm": 1.3043426791795303,
|
| 413 |
+
"learning_rate": 1.921012316518224e-05,
|
| 414 |
+
"loss": 1.2564,
|
| 415 |
+
"step": 570
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.46890000505280177,
|
| 419 |
+
"grad_norm": 1.4606599266501914,
|
| 420 |
+
"learning_rate": 1.917574110523994e-05,
|
| 421 |
+
"loss": 1.2455,
|
| 422 |
+
"step": 580
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.47698448789853976,
|
| 426 |
+
"grad_norm": 1.4084512281248918,
|
| 427 |
+
"learning_rate": 1.914065872168692e-05,
|
| 428 |
+
"loss": 1.237,
|
| 429 |
+
"step": 590
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.4850689707442777,
|
| 433 |
+
"grad_norm": 1.7442614101619536,
|
| 434 |
+
"learning_rate": 1.910487869212942e-05,
|
| 435 |
+
"loss": 1.2428,
|
| 436 |
+
"step": 600
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.49315345359001567,
|
| 440 |
+
"grad_norm": 3.217405569169141,
|
| 441 |
+
"learning_rate": 1.9068403747420365e-05,
|
| 442 |
+
"loss": 1.2406,
|
| 443 |
+
"step": 610
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5012379364357537,
|
| 447 |
+
"grad_norm": 2.028319411000034,
|
| 448 |
+
"learning_rate": 1.9031236671450963e-05,
|
| 449 |
+
"loss": 1.2295,
|
| 450 |
+
"step": 620
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5093224192814916,
|
| 454 |
+
"grad_norm": 1.3534016099705948,
|
| 455 |
+
"learning_rate": 1.899338030093822e-05,
|
| 456 |
+
"loss": 1.2287,
|
| 457 |
+
"step": 630
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5174069021272295,
|
| 461 |
+
"grad_norm": 16.560363141349324,
|
| 462 |
+
"learning_rate": 1.8954837525208432e-05,
|
| 463 |
+
"loss": 1.2239,
|
| 464 |
+
"step": 640
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5254913849729675,
|
| 468 |
+
"grad_norm": 1.6507428977016882,
|
| 469 |
+
"learning_rate": 1.8915611285976672e-05,
|
| 470 |
+
"loss": 1.2122,
|
| 471 |
+
"step": 650
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5335758678187055,
|
| 475 |
+
"grad_norm": 1.4599892786481696,
|
| 476 |
+
"learning_rate": 1.887570457712225e-05,
|
| 477 |
+
"loss": 1.2448,
|
| 478 |
+
"step": 660
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5416603506644434,
|
| 482 |
+
"grad_norm": 1.3576328654723246,
|
| 483 |
+
"learning_rate": 1.883512044446023e-05,
|
| 484 |
+
"loss": 1.235,
|
| 485 |
+
"step": 670
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5497448335101814,
|
| 489 |
+
"grad_norm": 2.7275818350144383,
|
| 490 |
+
"learning_rate": 1.879386198550895e-05,
|
| 491 |
+
"loss": 1.2302,
|
| 492 |
+
"step": 680
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5578293163559194,
|
| 496 |
+
"grad_norm": 1.4783688191374078,
|
| 497 |
+
"learning_rate": 1.8751932349253595e-05,
|
| 498 |
+
"loss": 1.2183,
|
| 499 |
+
"step": 690
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5659137992016573,
|
| 503 |
+
"grad_norm": 1.3696848099680126,
|
| 504 |
+
"learning_rate": 1.8709334735905908e-05,
|
| 505 |
+
"loss": 1.2202,
|
| 506 |
+
"step": 700
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5739982820473952,
|
| 510 |
+
"grad_norm": 1.3843064222445587,
|
| 511 |
+
"learning_rate": 1.866607239665988e-05,
|
| 512 |
+
"loss": 1.2292,
|
| 513 |
+
"step": 710
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5820827648931333,
|
| 517 |
+
"grad_norm": 1.3013446345815274,
|
| 518 |
+
"learning_rate": 1.8622148633443626e-05,
|
| 519 |
+
"loss": 1.2404,
|
| 520 |
+
"step": 720
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5901672477388712,
|
| 524 |
+
"grad_norm": 1.3389494076775972,
|
| 525 |
+
"learning_rate": 1.8577566798667397e-05,
|
| 526 |
+
"loss": 1.2,
|
| 527 |
+
"step": 730
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5982517305846091,
|
| 531 |
+
"grad_norm": 1.2803553653933784,
|
| 532 |
+
"learning_rate": 1.8532330294967678e-05,
|
| 533 |
+
"loss": 1.2019,
|
| 534 |
+
"step": 740
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6063362134303472,
|
| 538 |
+
"grad_norm": 1.3940783442430897,
|
| 539 |
+
"learning_rate": 1.848644257494751e-05,
|
| 540 |
+
"loss": 1.2111,
|
| 541 |
+
"step": 750
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6144206962760851,
|
| 545 |
+
"grad_norm": 1.2967372912925752,
|
| 546 |
+
"learning_rate": 1.8439907140912962e-05,
|
| 547 |
+
"loss": 1.2044,
|
| 548 |
+
"step": 760
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.622505179121823,
|
| 552 |
+
"grad_norm": 1.307050777866234,
|
| 553 |
+
"learning_rate": 1.839272754460583e-05,
|
| 554 |
+
"loss": 1.211,
|
| 555 |
+
"step": 770
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.630589661967561,
|
| 559 |
+
"grad_norm": 1.7851865803650349,
|
| 560 |
+
"learning_rate": 1.8344907386932552e-05,
|
| 561 |
+
"loss": 1.2038,
|
| 562 |
+
"step": 780
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.638674144813299,
|
| 566 |
+
"grad_norm": 1.8614266164299924,
|
| 567 |
+
"learning_rate": 1.8296450317689377e-05,
|
| 568 |
+
"loss": 1.2054,
|
| 569 |
+
"step": 790
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6467586276590369,
|
| 573 |
+
"grad_norm": 1.3262638540650757,
|
| 574 |
+
"learning_rate": 1.824736003528381e-05,
|
| 575 |
+
"loss": 1.209,
|
| 576 |
+
"step": 800
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.654843110504775,
|
| 580 |
+
"grad_norm": 1.290793353111858,
|
| 581 |
+
"learning_rate": 1.8197640286452312e-05,
|
| 582 |
+
"loss": 1.213,
|
| 583 |
+
"step": 810
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.6629275933505129,
|
| 587 |
+
"grad_norm": 1.2558226934999566,
|
| 588 |
+
"learning_rate": 1.814729486597436e-05,
|
| 589 |
+
"loss": 1.2266,
|
| 590 |
+
"step": 820
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6710120761962508,
|
| 594 |
+
"grad_norm": 1.277465841944589,
|
| 595 |
+
"learning_rate": 1.8096327616382815e-05,
|
| 596 |
+
"loss": 1.2167,
|
| 597 |
+
"step": 830
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6790965590419887,
|
| 601 |
+
"grad_norm": 1.298887855615747,
|
| 602 |
+
"learning_rate": 1.8044742427670627e-05,
|
| 603 |
+
"loss": 1.2226,
|
| 604 |
+
"step": 840
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6871810418877268,
|
| 608 |
+
"grad_norm": 5.857168222574854,
|
| 609 |
+
"learning_rate": 1.7992543236993952e-05,
|
| 610 |
+
"loss": 1.2027,
|
| 611 |
+
"step": 850
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.6952655247334647,
|
| 615 |
+
"grad_norm": 1.3361306728189393,
|
| 616 |
+
"learning_rate": 1.7939734028371663e-05,
|
| 617 |
+
"loss": 1.207,
|
| 618 |
+
"step": 860
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7033500075792026,
|
| 622 |
+
"grad_norm": 1.3969769044659528,
|
| 623 |
+
"learning_rate": 1.7886318832381264e-05,
|
| 624 |
+
"loss": 1.1799,
|
| 625 |
+
"step": 870
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7114344904249407,
|
| 629 |
+
"grad_norm": 1.4266930108547686,
|
| 630 |
+
"learning_rate": 1.783230172585126e-05,
|
| 631 |
+
"loss": 1.2111,
|
| 632 |
+
"step": 880
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7195189732706786,
|
| 636 |
+
"grad_norm": 1.3440902999919684,
|
| 637 |
+
"learning_rate": 1.7777686831550008e-05,
|
| 638 |
+
"loss": 1.1854,
|
| 639 |
+
"step": 890
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7276034561164165,
|
| 643 |
+
"grad_norm": 1.251718689797153,
|
| 644 |
+
"learning_rate": 1.7722478317871053e-05,
|
| 645 |
+
"loss": 1.1803,
|
| 646 |
+
"step": 900
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7356879389621546,
|
| 650 |
+
"grad_norm": 1.2756808323680056,
|
| 651 |
+
"learning_rate": 1.7666680398514978e-05,
|
| 652 |
+
"loss": 1.2148,
|
| 653 |
+
"step": 910
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.7437724218078925,
|
| 657 |
+
"grad_norm": 1.3774590120848857,
|
| 658 |
+
"learning_rate": 1.76102973321678e-05,
|
| 659 |
+
"loss": 1.189,
|
| 660 |
+
"step": 920
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7518569046536304,
|
| 664 |
+
"grad_norm": 1.5207360711907143,
|
| 665 |
+
"learning_rate": 1.7553333422175933e-05,
|
| 666 |
+
"loss": 1.1819,
|
| 667 |
+
"step": 930
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7599413874993683,
|
| 671 |
+
"grad_norm": 1.302009300658742,
|
| 672 |
+
"learning_rate": 1.7495793016217754e-05,
|
| 673 |
+
"loss": 1.191,
|
| 674 |
+
"step": 940
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7680258703451064,
|
| 678 |
+
"grad_norm": 1.3859954985668783,
|
| 679 |
+
"learning_rate": 1.743768050597175e-05,
|
| 680 |
+
"loss": 1.1835,
|
| 681 |
+
"step": 950
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7761103531908443,
|
| 685 |
+
"grad_norm": 1.3435502591474426,
|
| 686 |
+
"learning_rate": 1.7379000326781348e-05,
|
| 687 |
+
"loss": 1.2035,
|
| 688 |
+
"step": 960
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7841948360365822,
|
| 692 |
+
"grad_norm": 1.38981939520544,
|
| 693 |
+
"learning_rate": 1.7319756957316392e-05,
|
| 694 |
+
"loss": 1.1887,
|
| 695 |
+
"step": 970
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.7922793188823203,
|
| 699 |
+
"grad_norm": 1.4015519572670776,
|
| 700 |
+
"learning_rate": 1.725995491923131e-05,
|
| 701 |
+
"loss": 1.1843,
|
| 702 |
+
"step": 980
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8003638017280582,
|
| 706 |
+
"grad_norm": 1.4763071143801054,
|
| 707 |
+
"learning_rate": 1.7199598776820013e-05,
|
| 708 |
+
"loss": 1.1753,
|
| 709 |
+
"step": 990
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8084482845737961,
|
| 713 |
+
"grad_norm": 1.3577477544239007,
|
| 714 |
+
"learning_rate": 1.713869313666753e-05,
|
| 715 |
+
"loss": 1.1966,
|
| 716 |
+
"step": 1000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8165327674195342,
|
| 720 |
+
"grad_norm": 1.3963231420568032,
|
| 721 |
+
"learning_rate": 1.7077242647298405e-05,
|
| 722 |
+
"loss": 1.1985,
|
| 723 |
+
"step": 1010
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8246172502652721,
|
| 727 |
+
"grad_norm": 1.5498623314696613,
|
| 728 |
+
"learning_rate": 1.7015251998821938e-05,
|
| 729 |
+
"loss": 1.1785,
|
| 730 |
+
"step": 1020
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.83270173311101,
|
| 734 |
+
"grad_norm": 1.3586468512222978,
|
| 735 |
+
"learning_rate": 1.6952725922574188e-05,
|
| 736 |
+
"loss": 1.1648,
|
| 737 |
+
"step": 1030
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.840786215956748,
|
| 741 |
+
"grad_norm": 1.4300342736321576,
|
| 742 |
+
"learning_rate": 1.688966919075687e-05,
|
| 743 |
+
"loss": 1.1666,
|
| 744 |
+
"step": 1040
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.848870698802486,
|
| 748 |
+
"grad_norm": 1.5788283624417567,
|
| 749 |
+
"learning_rate": 1.682608661607313e-05,
|
| 750 |
+
"loss": 1.1821,
|
| 751 |
+
"step": 1050
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.8569551816482239,
|
| 755 |
+
"grad_norm": 1.359570582214726,
|
| 756 |
+
"learning_rate": 1.6761983051360232e-05,
|
| 757 |
+
"loss": 1.1958,
|
| 758 |
+
"step": 1060
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.8650396644939619,
|
| 762 |
+
"grad_norm": 1.3046392847858388,
|
| 763 |
+
"learning_rate": 1.6697363389219147e-05,
|
| 764 |
+
"loss": 1.1557,
|
| 765 |
+
"step": 1070
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.8731241473396999,
|
| 769 |
+
"grad_norm": 1.4677129965264875,
|
| 770 |
+
"learning_rate": 1.6632232561641158e-05,
|
| 771 |
+
"loss": 1.1593,
|
| 772 |
+
"step": 1080
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.8812086301854378,
|
| 776 |
+
"grad_norm": 1.4859252531152671,
|
| 777 |
+
"learning_rate": 1.6566595539631417e-05,
|
| 778 |
+
"loss": 1.1753,
|
| 779 |
+
"step": 1090
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8892931130311758,
|
| 783 |
+
"grad_norm": 1.3209365154297203,
|
| 784 |
+
"learning_rate": 1.6500457332829553e-05,
|
| 785 |
+
"loss": 1.161,
|
| 786 |
+
"step": 1100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.8973775958769138,
|
| 790 |
+
"grad_norm": 1.3862159117294945,
|
| 791 |
+
"learning_rate": 1.6433822989127314e-05,
|
| 792 |
+
"loss": 1.1592,
|
| 793 |
+
"step": 1110
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9054620787226517,
|
| 797 |
+
"grad_norm": 1.4456179949854164,
|
| 798 |
+
"learning_rate": 1.636669759428329e-05,
|
| 799 |
+
"loss": 1.1484,
|
| 800 |
+
"step": 1120
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9135465615683896,
|
| 804 |
+
"grad_norm": 1.288756152636894,
|
| 805 |
+
"learning_rate": 1.6299086271534764e-05,
|
| 806 |
+
"loss": 1.181,
|
| 807 |
+
"step": 1130
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.9216310444141277,
|
| 811 |
+
"grad_norm": 1.2599229391965052,
|
| 812 |
+
"learning_rate": 1.6230994181206674e-05,
|
| 813 |
+
"loss": 1.1718,
|
| 814 |
+
"step": 1140
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.9297155272598656,
|
| 818 |
+
"grad_norm": 1.4973902946133841,
|
| 819 |
+
"learning_rate": 1.6162426520317765e-05,
|
| 820 |
+
"loss": 1.1773,
|
| 821 |
+
"step": 1150
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.9378000101056035,
|
| 825 |
+
"grad_norm": 1.3698767908727083,
|
| 826 |
+
"learning_rate": 1.6093388522183948e-05,
|
| 827 |
+
"loss": 1.1666,
|
| 828 |
+
"step": 1160
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.9458844929513415,
|
| 832 |
+
"grad_norm": 1.386433062647111,
|
| 833 |
+
"learning_rate": 1.6023885456018852e-05,
|
| 834 |
+
"loss": 1.1859,
|
| 835 |
+
"step": 1170
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9539689757970795,
|
| 839 |
+
"grad_norm": 1.284904254015402,
|
| 840 |
+
"learning_rate": 1.595392262653168e-05,
|
| 841 |
+
"loss": 1.1906,
|
| 842 |
+
"step": 1180
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.9620534586428174,
|
| 846 |
+
"grad_norm": 1.4402131637475677,
|
| 847 |
+
"learning_rate": 1.5883505373522317e-05,
|
| 848 |
+
"loss": 1.1593,
|
| 849 |
+
"step": 1190
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.9701379414885554,
|
| 853 |
+
"grad_norm": 1.6049356540049453,
|
| 854 |
+
"learning_rate": 1.5812639071473804e-05,
|
| 855 |
+
"loss": 1.1636,
|
| 856 |
+
"step": 1200
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.9782224243342934,
|
| 860 |
+
"grad_norm": 1.505036374645861,
|
| 861 |
+
"learning_rate": 1.574132912914211e-05,
|
| 862 |
+
"loss": 1.14,
|
| 863 |
+
"step": 1210
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.9863069071800313,
|
| 867 |
+
"grad_norm": 1.6280895974825729,
|
| 868 |
+
"learning_rate": 1.566958098914334e-05,
|
| 869 |
+
"loss": 1.1358,
|
| 870 |
+
"step": 1220
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.9943913900257693,
|
| 874 |
+
"grad_norm": 1.2574161457807662,
|
| 875 |
+
"learning_rate": 1.5597400127538324e-05,
|
| 876 |
+
"loss": 1.1754,
|
| 877 |
+
"step": 1230
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.9992420797332121,
|
| 881 |
+
"eval_loss": 1.0555766820907593,
|
| 882 |
+
"eval_runtime": 476.758,
|
| 883 |
+
"eval_samples_per_second": 25.514,
|
| 884 |
+
"eval_steps_per_second": 12.757,
|
| 885 |
+
"step": 1236
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0024758728715073,
|
| 889 |
+
"grad_norm": 2.9356360899500897,
|
| 890 |
+
"learning_rate": 1.5524792053414676e-05,
|
| 891 |
+
"loss": 1.1182,
|
| 892 |
+
"step": 1240
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0105603557172451,
|
| 896 |
+
"grad_norm": 1.4115997260524025,
|
| 897 |
+
"learning_rate": 1.5451762308466302e-05,
|
| 898 |
+
"loss": 1.0448,
|
| 899 |
+
"step": 1250
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0186448385629832,
|
| 903 |
+
"grad_norm": 1.4408354404654395,
|
| 904 |
+
"learning_rate": 1.5378316466570466e-05,
|
| 905 |
+
"loss": 1.027,
|
| 906 |
+
"step": 1260
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0267293214087212,
|
| 910 |
+
"grad_norm": 1.40209737150782,
|
| 911 |
+
"learning_rate": 1.530446013336235e-05,
|
| 912 |
+
"loss": 1.0253,
|
| 913 |
+
"step": 1270
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.034813804254459,
|
| 917 |
+
"grad_norm": 1.4050923085204698,
|
| 918 |
+
"learning_rate": 1.5230198945807226e-05,
|
| 919 |
+
"loss": 1.0596,
|
| 920 |
+
"step": 1280
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.042898287100197,
|
| 924 |
+
"grad_norm": 1.3850604464116953,
|
| 925 |
+
"learning_rate": 1.515553857177022e-05,
|
| 926 |
+
"loss": 1.0354,
|
| 927 |
+
"step": 1290
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.050982769945935,
|
| 931 |
+
"grad_norm": 1.6192982769908866,
|
| 932 |
+
"learning_rate": 1.5080484709583715e-05,
|
| 933 |
+
"loss": 1.0338,
|
| 934 |
+
"step": 1300
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.059067252791673,
|
| 938 |
+
"grad_norm": 1.5443333411983042,
|
| 939 |
+
"learning_rate": 1.5005043087612452e-05,
|
| 940 |
+
"loss": 1.0224,
|
| 941 |
+
"step": 1310
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.067151735637411,
|
| 945 |
+
"grad_norm": 1.4795375887873081,
|
| 946 |
+
"learning_rate": 1.4929219463816302e-05,
|
| 947 |
+
"loss": 1.0273,
|
| 948 |
+
"step": 1320
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.075236218483149,
|
| 952 |
+
"grad_norm": 1.3952469643942318,
|
| 953 |
+
"learning_rate": 1.4853019625310813e-05,
|
| 954 |
+
"loss": 1.0165,
|
| 955 |
+
"step": 1330
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.0833207013288868,
|
| 959 |
+
"grad_norm": 1.4102438583126526,
|
| 960 |
+
"learning_rate": 1.4776449387925507e-05,
|
| 961 |
+
"loss": 1.0323,
|
| 962 |
+
"step": 1340
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.0914051841746248,
|
| 966 |
+
"grad_norm": 1.4166513317270177,
|
| 967 |
+
"learning_rate": 1.4699514595760006e-05,
|
| 968 |
+
"loss": 1.0343,
|
| 969 |
+
"step": 1350
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.0994896670203629,
|
| 973 |
+
"grad_norm": 1.4572773218335806,
|
| 974 |
+
"learning_rate": 1.4622221120737985e-05,
|
| 975 |
+
"loss": 1.0449,
|
| 976 |
+
"step": 1360
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.1075741498661007,
|
| 980 |
+
"grad_norm": 1.4277575864922984,
|
| 981 |
+
"learning_rate": 1.4544574862159013e-05,
|
| 982 |
+
"loss": 1.0157,
|
| 983 |
+
"step": 1370
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1156586327118387,
|
| 987 |
+
"grad_norm": 1.8246683293221693,
|
| 988 |
+
"learning_rate": 1.446658174624829e-05,
|
| 989 |
+
"loss": 1.037,
|
| 990 |
+
"step": 1380
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1237431155575768,
|
| 994 |
+
"grad_norm": 1.4515508954548648,
|
| 995 |
+
"learning_rate": 1.4388247725704338e-05,
|
| 996 |
+
"loss": 1.0163,
|
| 997 |
+
"step": 1390
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1318275984033146,
|
| 1001 |
+
"grad_norm": 1.4472625641065484,
|
| 1002 |
+
"learning_rate": 1.4309578779244678e-05,
|
| 1003 |
+
"loss": 1.0339,
|
| 1004 |
+
"step": 1400
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1399120812490526,
|
| 1008 |
+
"grad_norm": 1.441284439472294,
|
| 1009 |
+
"learning_rate": 1.423058091114951e-05,
|
| 1010 |
+
"loss": 1.0153,
|
| 1011 |
+
"step": 1410
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.1479965640947905,
|
| 1015 |
+
"grad_norm": 1.4505444065925723,
|
| 1016 |
+
"learning_rate": 1.4151260150803445e-05,
|
| 1017 |
+
"loss": 1.0413,
|
| 1018 |
+
"step": 1420
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.1560810469405285,
|
| 1022 |
+
"grad_norm": 1.5566575848024742,
|
| 1023 |
+
"learning_rate": 1.4071622552235327e-05,
|
| 1024 |
+
"loss": 1.014,
|
| 1025 |
+
"step": 1430
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.1641655297862665,
|
| 1029 |
+
"grad_norm": 1.476527456836737,
|
| 1030 |
+
"learning_rate": 1.399167419365616e-05,
|
| 1031 |
+
"loss": 1.0374,
|
| 1032 |
+
"step": 1440
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.1722500126320043,
|
| 1036 |
+
"grad_norm": 1.7587555981022083,
|
| 1037 |
+
"learning_rate": 1.3911421176995206e-05,
|
| 1038 |
+
"loss": 1.0145,
|
| 1039 |
+
"step": 1450
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.1803344954777424,
|
| 1043 |
+
"grad_norm": 1.5447530212974045,
|
| 1044 |
+
"learning_rate": 1.3830869627434267e-05,
|
| 1045 |
+
"loss": 1.0104,
|
| 1046 |
+
"step": 1460
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.1884189783234804,
|
| 1050 |
+
"grad_norm": 1.368002967716879,
|
| 1051 |
+
"learning_rate": 1.3750025692940174e-05,
|
| 1052 |
+
"loss": 1.0102,
|
| 1053 |
+
"step": 1470
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.1965034611692182,
|
| 1057 |
+
"grad_norm": 1.5132346329088506,
|
| 1058 |
+
"learning_rate": 1.3668895543795581e-05,
|
| 1059 |
+
"loss": 1.0241,
|
| 1060 |
+
"step": 1480
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.2045879440149563,
|
| 1064 |
+
"grad_norm": 1.4535090384504317,
|
| 1065 |
+
"learning_rate": 1.3587485372128e-05,
|
| 1066 |
+
"loss": 1.01,
|
| 1067 |
+
"step": 1490
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2126724268606943,
|
| 1071 |
+
"grad_norm": 1.6349536867702466,
|
| 1072 |
+
"learning_rate": 1.3505801391437215e-05,
|
| 1073 |
+
"loss": 1.0538,
|
| 1074 |
+
"step": 1500
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.2207569097064321,
|
| 1078 |
+
"grad_norm": 1.608679365926187,
|
| 1079 |
+
"learning_rate": 1.3423849836121043e-05,
|
| 1080 |
+
"loss": 1.0256,
|
| 1081 |
+
"step": 1510
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2288413925521702,
|
| 1085 |
+
"grad_norm": 1.4875509565909706,
|
| 1086 |
+
"learning_rate": 1.33416369609995e-05,
|
| 1087 |
+
"loss": 1.0365,
|
| 1088 |
+
"step": 1520
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2369258753979082,
|
| 1092 |
+
"grad_norm": 1.4161399144655036,
|
| 1093 |
+
"learning_rate": 1.325916904083741e-05,
|
| 1094 |
+
"loss": 1.0285,
|
| 1095 |
+
"step": 1530
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.245010358243646,
|
| 1099 |
+
"grad_norm": 1.516547180031239,
|
| 1100 |
+
"learning_rate": 1.3176452369865504e-05,
|
| 1101 |
+
"loss": 0.9972,
|
| 1102 |
+
"step": 1540
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.253094841089384,
|
| 1106 |
+
"grad_norm": 1.4500310981963098,
|
| 1107 |
+
"learning_rate": 1.3093493261300012e-05,
|
| 1108 |
+
"loss": 1.0122,
|
| 1109 |
+
"step": 1550
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.261179323935122,
|
| 1113 |
+
"grad_norm": 1.3787551364346502,
|
| 1114 |
+
"learning_rate": 1.3010298046860821e-05,
|
| 1115 |
+
"loss": 1.0221,
|
| 1116 |
+
"step": 1560
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.26926380678086,
|
| 1120 |
+
"grad_norm": 1.3579456863416077,
|
| 1121 |
+
"learning_rate": 1.2926873076288222e-05,
|
| 1122 |
+
"loss": 1.0213,
|
| 1123 |
+
"step": 1570
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.277348289626598,
|
| 1127 |
+
"grad_norm": 1.4774509503134268,
|
| 1128 |
+
"learning_rate": 1.2843224716858271e-05,
|
| 1129 |
+
"loss": 1.012,
|
| 1130 |
+
"step": 1580
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.285432772472336,
|
| 1134 |
+
"grad_norm": 1.4805342986177266,
|
| 1135 |
+
"learning_rate": 1.2759359352896809e-05,
|
| 1136 |
+
"loss": 1.0193,
|
| 1137 |
+
"step": 1590
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.2935172553180738,
|
| 1141 |
+
"grad_norm": 1.4527468028008124,
|
| 1142 |
+
"learning_rate": 1.2675283385292212e-05,
|
| 1143 |
+
"loss": 1.0431,
|
| 1144 |
+
"step": 1600
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.3016017381638119,
|
| 1148 |
+
"grad_norm": 1.5688075844044822,
|
| 1149 |
+
"learning_rate": 1.259100323100682e-05,
|
| 1150 |
+
"loss": 1.0226,
|
| 1151 |
+
"step": 1610
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.30968622100955,
|
| 1155 |
+
"grad_norm": 1.493324687221304,
|
| 1156 |
+
"learning_rate": 1.2506525322587207e-05,
|
| 1157 |
+
"loss": 0.9966,
|
| 1158 |
+
"step": 1620
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.3177707038552877,
|
| 1162 |
+
"grad_norm": 1.563824009098089,
|
| 1163 |
+
"learning_rate": 1.2421856107673205e-05,
|
| 1164 |
+
"loss": 1.0317,
|
| 1165 |
+
"step": 1630
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3258551867010258,
|
| 1169 |
+
"grad_norm": 1.4698666764020467,
|
| 1170 |
+
"learning_rate": 1.233700204850581e-05,
|
| 1171 |
+
"loss": 1.0013,
|
| 1172 |
+
"step": 1640
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.3339396695467638,
|
| 1176 |
+
"grad_norm": 1.625463847709757,
|
| 1177 |
+
"learning_rate": 1.2251969621433947e-05,
|
| 1178 |
+
"loss": 1.0233,
|
| 1179 |
+
"step": 1650
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.3420241523925016,
|
| 1183 |
+
"grad_norm": 1.560576858468798,
|
| 1184 |
+
"learning_rate": 1.2166765316420195e-05,
|
| 1185 |
+
"loss": 1.0137,
|
| 1186 |
+
"step": 1660
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.3501086352382397,
|
| 1190 |
+
"grad_norm": 1.6305115869655395,
|
| 1191 |
+
"learning_rate": 1.2081395636545432e-05,
|
| 1192 |
+
"loss": 1.0074,
|
| 1193 |
+
"step": 1670
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.3581931180839777,
|
| 1197 |
+
"grad_norm": 1.683367869903662,
|
| 1198 |
+
"learning_rate": 1.1995867097512504e-05,
|
| 1199 |
+
"loss": 1.0202,
|
| 1200 |
+
"step": 1680
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.3662776009297155,
|
| 1204 |
+
"grad_norm": 1.342629975477622,
|
| 1205 |
+
"learning_rate": 1.191018622714893e-05,
|
| 1206 |
+
"loss": 1.0039,
|
| 1207 |
+
"step": 1690
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.3743620837754535,
|
| 1211 |
+
"grad_norm": 1.4162506108365653,
|
| 1212 |
+
"learning_rate": 1.1824359564908667e-05,
|
| 1213 |
+
"loss": 1.0303,
|
| 1214 |
+
"step": 1700
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.3824465666211916,
|
| 1218 |
+
"grad_norm": 1.4322509952288762,
|
| 1219 |
+
"learning_rate": 1.1738393661373004e-05,
|
| 1220 |
+
"loss": 1.0223,
|
| 1221 |
+
"step": 1710
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.3905310494669294,
|
| 1225 |
+
"grad_norm": 1.4429525488762647,
|
| 1226 |
+
"learning_rate": 1.1652295077750599e-05,
|
| 1227 |
+
"loss": 1.0079,
|
| 1228 |
+
"step": 1720
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.3986155323126674,
|
| 1232 |
+
"grad_norm": 1.5044521870868257,
|
| 1233 |
+
"learning_rate": 1.1566070385376705e-05,
|
| 1234 |
+
"loss": 0.9903,
|
| 1235 |
+
"step": 1730
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.4067000151584053,
|
| 1239 |
+
"grad_norm": 1.4591518605463256,
|
| 1240 |
+
"learning_rate": 1.1479726165211609e-05,
|
| 1241 |
+
"loss": 1.0133,
|
| 1242 |
+
"step": 1740
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4147844980041433,
|
| 1246 |
+
"grad_norm": 1.38699009818023,
|
| 1247 |
+
"learning_rate": 1.1393269007338375e-05,
|
| 1248 |
+
"loss": 1.0191,
|
| 1249 |
+
"step": 1750
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.4228689808498813,
|
| 1253 |
+
"grad_norm": 1.4248174199771946,
|
| 1254 |
+
"learning_rate": 1.1306705510459852e-05,
|
| 1255 |
+
"loss": 1.0048,
|
| 1256 |
+
"step": 1760
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.4309534636956192,
|
| 1260 |
+
"grad_norm": 1.5368128288739022,
|
| 1261 |
+
"learning_rate": 1.1220042281395042e-05,
|
| 1262 |
+
"loss": 1.0169,
|
| 1263 |
+
"step": 1770
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4390379465413572,
|
| 1267 |
+
"grad_norm": 1.620365193180215,
|
| 1268 |
+
"learning_rate": 1.1133285934574849e-05,
|
| 1269 |
+
"loss": 0.9982,
|
| 1270 |
+
"step": 1780
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.447122429387095,
|
| 1274 |
+
"grad_norm": 1.4821421519804139,
|
| 1275 |
+
"learning_rate": 1.1046443091537232e-05,
|
| 1276 |
+
"loss": 1.0241,
|
| 1277 |
+
"step": 1790
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.455206912232833,
|
| 1281 |
+
"grad_norm": 1.5012997646705204,
|
| 1282 |
+
"learning_rate": 1.0959520380421831e-05,
|
| 1283 |
+
"loss": 1.0116,
|
| 1284 |
+
"step": 1800
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.463291395078571,
|
| 1288 |
+
"grad_norm": 1.4878335919543981,
|
| 1289 |
+
"learning_rate": 1.0872524435464104e-05,
|
| 1290 |
+
"loss": 0.9993,
|
| 1291 |
+
"step": 1810
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.471375877924309,
|
| 1295 |
+
"grad_norm": 1.3918759318142178,
|
| 1296 |
+
"learning_rate": 1.0785461896488947e-05,
|
| 1297 |
+
"loss": 1.0103,
|
| 1298 |
+
"step": 1820
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.479460360770047,
|
| 1302 |
+
"grad_norm": 1.7724767013914755,
|
| 1303 |
+
"learning_rate": 1.0698339408403944e-05,
|
| 1304 |
+
"loss": 0.9862,
|
| 1305 |
+
"step": 1830
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.487544843615785,
|
| 1309 |
+
"grad_norm": 2.0093844914876717,
|
| 1310 |
+
"learning_rate": 1.06111636206922e-05,
|
| 1311 |
+
"loss": 1.0039,
|
| 1312 |
+
"step": 1840
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.4956293264615228,
|
| 1316 |
+
"grad_norm": 1.4440349729006745,
|
| 1317 |
+
"learning_rate": 1.0523941186904823e-05,
|
| 1318 |
+
"loss": 1.0091,
|
| 1319 |
+
"step": 1850
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.5037138093072608,
|
| 1323 |
+
"grad_norm": 1.5530469064140777,
|
| 1324 |
+
"learning_rate": 1.043667876415311e-05,
|
| 1325 |
+
"loss": 0.9959,
|
| 1326 |
+
"step": 1860
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.5117982921529989,
|
| 1330 |
+
"grad_norm": 1.9710010624543786,
|
| 1331 |
+
"learning_rate": 1.0349383012600448e-05,
|
| 1332 |
+
"loss": 0.9902,
|
| 1333 |
+
"step": 1870
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.5198827749987367,
|
| 1337 |
+
"grad_norm": 1.4874119470603941,
|
| 1338 |
+
"learning_rate": 1.0262060594954e-05,
|
| 1339 |
+
"loss": 0.9889,
|
| 1340 |
+
"step": 1880
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.5279672578444747,
|
| 1344 |
+
"grad_norm": 1.5760932908781828,
|
| 1345 |
+
"learning_rate": 1.0174718175956164e-05,
|
| 1346 |
+
"loss": 0.997,
|
| 1347 |
+
"step": 1890
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.5360517406902128,
|
| 1351 |
+
"grad_norm": 1.5140336706570001,
|
| 1352 |
+
"learning_rate": 1.0087362421875912e-05,
|
| 1353 |
+
"loss": 1.0162,
|
| 1354 |
+
"step": 1900
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.5441362235359506,
|
| 1358 |
+
"grad_norm": 1.4275012742483075,
|
| 1359 |
+
"learning_rate": 1e-05,
|
| 1360 |
+
"loss": 1.0056,
|
| 1361 |
+
"step": 1910
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.5522207063816886,
|
| 1365 |
+
"grad_norm": 1.4479646715349155,
|
| 1366 |
+
"learning_rate": 9.912637578124092e-06,
|
| 1367 |
+
"loss": 0.9831,
|
| 1368 |
+
"step": 1920
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.5603051892274267,
|
| 1372 |
+
"grad_norm": 1.6529106306573094,
|
| 1373 |
+
"learning_rate": 9.825281824043838e-06,
|
| 1374 |
+
"loss": 1.0009,
|
| 1375 |
+
"step": 1930
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.5683896720731645,
|
| 1379 |
+
"grad_norm": 1.4537655155385498,
|
| 1380 |
+
"learning_rate": 9.737939405046002e-06,
|
| 1381 |
+
"loss": 1.0058,
|
| 1382 |
+
"step": 1940
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.5764741549189025,
|
| 1386 |
+
"grad_norm": 1.3881828231981752,
|
| 1387 |
+
"learning_rate": 9.650616987399553e-06,
|
| 1388 |
+
"loss": 0.9752,
|
| 1389 |
+
"step": 1950
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.5845586377646406,
|
| 1393 |
+
"grad_norm": 1.4410127433172688,
|
| 1394 |
+
"learning_rate": 9.563321235846894e-06,
|
| 1395 |
+
"loss": 1.0026,
|
| 1396 |
+
"step": 1960
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.5926431206103784,
|
| 1400 |
+
"grad_norm": 1.6585729752037028,
|
| 1401 |
+
"learning_rate": 9.476058813095182e-06,
|
| 1402 |
+
"loss": 0.9942,
|
| 1403 |
+
"step": 1970
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.6007276034561164,
|
| 1407 |
+
"grad_norm": 1.6572316797520206,
|
| 1408 |
+
"learning_rate": 9.388836379307802e-06,
|
| 1409 |
+
"loss": 0.9968,
|
| 1410 |
+
"step": 1980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.6088120863018545,
|
| 1414 |
+
"grad_norm": 1.451151024162774,
|
| 1415 |
+
"learning_rate": 9.301660591596059e-06,
|
| 1416 |
+
"loss": 0.9921,
|
| 1417 |
+
"step": 1990
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.6168965691475923,
|
| 1421 |
+
"grad_norm": 1.5042478185497792,
|
| 1422 |
+
"learning_rate": 9.214538103511053e-06,
|
| 1423 |
+
"loss": 0.9959,
|
| 1424 |
+
"step": 2000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6249810519933303,
|
| 1428 |
+
"grad_norm": 1.4096442655309245,
|
| 1429 |
+
"learning_rate": 9.127475564535898e-06,
|
| 1430 |
+
"loss": 0.9944,
|
| 1431 |
+
"step": 2010
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.6330655348390684,
|
| 1435 |
+
"grad_norm": 1.3701103693221475,
|
| 1436 |
+
"learning_rate": 9.04047961957817e-06,
|
| 1437 |
+
"loss": 0.9806,
|
| 1438 |
+
"step": 2020
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.6411500176848062,
|
| 1442 |
+
"grad_norm": 1.6771886101217564,
|
| 1443 |
+
"learning_rate": 8.953556908462773e-06,
|
| 1444 |
+
"loss": 0.9986,
|
| 1445 |
+
"step": 2030
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.6492345005305442,
|
| 1449 |
+
"grad_norm": 1.4606744478213272,
|
| 1450 |
+
"learning_rate": 8.866714065425154e-06,
|
| 1451 |
+
"loss": 0.9894,
|
| 1452 |
+
"step": 2040
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.6573189833762823,
|
| 1456 |
+
"grad_norm": 1.5696191298486186,
|
| 1457 |
+
"learning_rate": 8.779957718604956e-06,
|
| 1458 |
+
"loss": 1.0055,
|
| 1459 |
+
"step": 2050
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.66540346622202,
|
| 1463 |
+
"grad_norm": 1.4621439613400917,
|
| 1464 |
+
"learning_rate": 8.693294489540151e-06,
|
| 1465 |
+
"loss": 1.0055,
|
| 1466 |
+
"step": 2060
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.673487949067758,
|
| 1470 |
+
"grad_norm": 1.4224764910826249,
|
| 1471 |
+
"learning_rate": 8.60673099266163e-06,
|
| 1472 |
+
"loss": 0.9687,
|
| 1473 |
+
"step": 2070
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.6815724319134961,
|
| 1477 |
+
"grad_norm": 1.6938323822086323,
|
| 1478 |
+
"learning_rate": 8.520273834788395e-06,
|
| 1479 |
+
"loss": 0.978,
|
| 1480 |
+
"step": 2080
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.689656914759234,
|
| 1484 |
+
"grad_norm": 1.5856717495753165,
|
| 1485 |
+
"learning_rate": 8.4339296146233e-06,
|
| 1486 |
+
"loss": 0.992,
|
| 1487 |
+
"step": 2090
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.697741397604972,
|
| 1491 |
+
"grad_norm": 1.4737528022353619,
|
| 1492 |
+
"learning_rate": 8.3477049222494e-06,
|
| 1493 |
+
"loss": 0.9882,
|
| 1494 |
+
"step": 2100
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.70582588045071,
|
| 1498 |
+
"grad_norm": 1.4413576604331515,
|
| 1499 |
+
"learning_rate": 8.261606338626998e-06,
|
| 1500 |
+
"loss": 0.9717,
|
| 1501 |
+
"step": 2110
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.7139103632964479,
|
| 1505 |
+
"grad_norm": 1.4533604100239785,
|
| 1506 |
+
"learning_rate": 8.17564043509134e-06,
|
| 1507 |
+
"loss": 0.9878,
|
| 1508 |
+
"step": 2120
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.7219948461421857,
|
| 1512 |
+
"grad_norm": 1.4996211527080612,
|
| 1513 |
+
"learning_rate": 8.089813772851073e-06,
|
| 1514 |
+
"loss": 0.9932,
|
| 1515 |
+
"step": 2130
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.730079328987924,
|
| 1519 |
+
"grad_norm": 1.4183735479797297,
|
| 1520 |
+
"learning_rate": 8.004132902487499e-06,
|
| 1521 |
+
"loss": 1.0021,
|
| 1522 |
+
"step": 2140
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.7381638118336618,
|
| 1526 |
+
"grad_norm": 1.4020103234354604,
|
| 1527 |
+
"learning_rate": 7.91860436345457e-06,
|
| 1528 |
+
"loss": 0.9717,
|
| 1529 |
+
"step": 2150
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.7462482946793996,
|
| 1533 |
+
"grad_norm": 1.4529101522297827,
|
| 1534 |
+
"learning_rate": 7.833234683579806e-06,
|
| 1535 |
+
"loss": 0.9844,
|
| 1536 |
+
"step": 2160
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.7543327775251378,
|
| 1540 |
+
"grad_norm": 1.4502465958251158,
|
| 1541 |
+
"learning_rate": 7.748030378566056e-06,
|
| 1542 |
+
"loss": 0.9782,
|
| 1543 |
+
"step": 2170
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.7624172603708756,
|
| 1547 |
+
"grad_norm": 1.4461707858445054,
|
| 1548 |
+
"learning_rate": 7.662997951494193e-06,
|
| 1549 |
+
"loss": 0.9836,
|
| 1550 |
+
"step": 2180
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.7705017432166135,
|
| 1554 |
+
"grad_norm": 1.3966480403360386,
|
| 1555 |
+
"learning_rate": 7.578143892326797e-06,
|
| 1556 |
+
"loss": 1.0089,
|
| 1557 |
+
"step": 2190
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.7785862260623517,
|
| 1561 |
+
"grad_norm": 1.5838575969719086,
|
| 1562 |
+
"learning_rate": 7.493474677412795e-06,
|
| 1563 |
+
"loss": 1.0017,
|
| 1564 |
+
"step": 2200
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.7866707089080895,
|
| 1568 |
+
"grad_norm": 1.6412461821364432,
|
| 1569 |
+
"learning_rate": 7.408996768993184e-06,
|
| 1570 |
+
"loss": 0.9889,
|
| 1571 |
+
"step": 2210
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.7947551917538274,
|
| 1575 |
+
"grad_norm": 1.8686882471940454,
|
| 1576 |
+
"learning_rate": 7.324716614707794e-06,
|
| 1577 |
+
"loss": 0.9814,
|
| 1578 |
+
"step": 2220
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.8028396745995656,
|
| 1582 |
+
"grad_norm": 1.4444454657231485,
|
| 1583 |
+
"learning_rate": 7.240640647103192e-06,
|
| 1584 |
+
"loss": 0.9934,
|
| 1585 |
+
"step": 2230
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.8109241574453034,
|
| 1589 |
+
"grad_norm": 1.5880994051473134,
|
| 1590 |
+
"learning_rate": 7.156775283141733e-06,
|
| 1591 |
+
"loss": 0.9972,
|
| 1592 |
+
"step": 2240
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.8190086402910413,
|
| 1596 |
+
"grad_norm": 1.6179768250952558,
|
| 1597 |
+
"learning_rate": 7.0731269237117775e-06,
|
| 1598 |
+
"loss": 0.9805,
|
| 1599 |
+
"step": 2250
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.8270931231367793,
|
| 1603 |
+
"grad_norm": 1.4161571668846493,
|
| 1604 |
+
"learning_rate": 6.989701953139181e-06,
|
| 1605 |
+
"loss": 0.9695,
|
| 1606 |
+
"step": 2260
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.8351776059825173,
|
| 1610 |
+
"grad_norm": 1.8752619329260358,
|
| 1611 |
+
"learning_rate": 6.906506738699994e-06,
|
| 1612 |
+
"loss": 0.9899,
|
| 1613 |
+
"step": 2270
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.8432620888282552,
|
| 1617 |
+
"grad_norm": 1.8476640791436918,
|
| 1618 |
+
"learning_rate": 6.823547630134497e-06,
|
| 1619 |
+
"loss": 0.9799,
|
| 1620 |
+
"step": 2280
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.8513465716739932,
|
| 1624 |
+
"grad_norm": 1.5003229948984453,
|
| 1625 |
+
"learning_rate": 6.740830959162592e-06,
|
| 1626 |
+
"loss": 0.9948,
|
| 1627 |
+
"step": 2290
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.8594310545197312,
|
| 1631 |
+
"grad_norm": 1.4363919724793655,
|
| 1632 |
+
"learning_rate": 6.658363039000501e-06,
|
| 1633 |
+
"loss": 0.9625,
|
| 1634 |
+
"step": 2300
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.867515537365469,
|
| 1638 |
+
"grad_norm": 1.45857815520064,
|
| 1639 |
+
"learning_rate": 6.57615016387896e-06,
|
| 1640 |
+
"loss": 0.976,
|
| 1641 |
+
"step": 2310
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.875600020211207,
|
| 1645 |
+
"grad_norm": 1.3637017381911254,
|
| 1646 |
+
"learning_rate": 6.4941986085627895e-06,
|
| 1647 |
+
"loss": 0.9608,
|
| 1648 |
+
"step": 2320
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.8836845030569451,
|
| 1652 |
+
"grad_norm": 1.586134857640991,
|
| 1653 |
+
"learning_rate": 6.412514627872003e-06,
|
| 1654 |
+
"loss": 0.9702,
|
| 1655 |
+
"step": 2330
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.891768985902683,
|
| 1659 |
+
"grad_norm": 1.6293874205755696,
|
| 1660 |
+
"learning_rate": 6.331104456204423e-06,
|
| 1661 |
+
"loss": 0.9672,
|
| 1662 |
+
"step": 2340
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.899853468748421,
|
| 1666 |
+
"grad_norm": 1.6185456719315228,
|
| 1667 |
+
"learning_rate": 6.249974307059826e-06,
|
| 1668 |
+
"loss": 0.9683,
|
| 1669 |
+
"step": 2350
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.907937951594159,
|
| 1673 |
+
"grad_norm": 1.5897776438113254,
|
| 1674 |
+
"learning_rate": 6.169130372565737e-06,
|
| 1675 |
+
"loss": 0.9942,
|
| 1676 |
+
"step": 2360
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.9160224344398968,
|
| 1680 |
+
"grad_norm": 1.4621464766459995,
|
| 1681 |
+
"learning_rate": 6.088578823004796e-06,
|
| 1682 |
+
"loss": 0.9552,
|
| 1683 |
+
"step": 2370
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.9241069172856349,
|
| 1687 |
+
"grad_norm": 1.57419066036152,
|
| 1688 |
+
"learning_rate": 6.008325806343842e-06,
|
| 1689 |
+
"loss": 0.9635,
|
| 1690 |
+
"step": 2380
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.932191400131373,
|
| 1694 |
+
"grad_norm": 1.4154240767952921,
|
| 1695 |
+
"learning_rate": 5.9283774477646775e-06,
|
| 1696 |
+
"loss": 0.9661,
|
| 1697 |
+
"step": 2390
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.9402758829771107,
|
| 1701 |
+
"grad_norm": 1.4089774352311322,
|
| 1702 |
+
"learning_rate": 5.848739849196556e-06,
|
| 1703 |
+
"loss": 0.9623,
|
| 1704 |
+
"step": 2400
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.9483603658228488,
|
| 1708 |
+
"grad_norm": 1.4330997113061938,
|
| 1709 |
+
"learning_rate": 5.7694190888504964e-06,
|
| 1710 |
+
"loss": 0.982,
|
| 1711 |
+
"step": 2410
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.9564448486685868,
|
| 1715 |
+
"grad_norm": 1.762833270995275,
|
| 1716 |
+
"learning_rate": 5.690421220755329e-06,
|
| 1717 |
+
"loss": 0.968,
|
| 1718 |
+
"step": 2420
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.9645293315143246,
|
| 1722 |
+
"grad_norm": 1.57370551896378,
|
| 1723 |
+
"learning_rate": 5.611752274295665e-06,
|
| 1724 |
+
"loss": 0.9639,
|
| 1725 |
+
"step": 2430
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.9726138143600627,
|
| 1729 |
+
"grad_norm": 1.4682932578058885,
|
| 1730 |
+
"learning_rate": 5.533418253751714e-06,
|
| 1731 |
+
"loss": 0.9786,
|
| 1732 |
+
"step": 2440
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.9806982972058007,
|
| 1736 |
+
"grad_norm": 1.7633821953728437,
|
| 1737 |
+
"learning_rate": 5.455425137840987e-06,
|
| 1738 |
+
"loss": 0.9618,
|
| 1739 |
+
"step": 2450
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.9887827800515385,
|
| 1743 |
+
"grad_norm": 1.5018261369656176,
|
| 1744 |
+
"learning_rate": 5.377778879262017e-06,
|
| 1745 |
+
"loss": 0.9454,
|
| 1746 |
+
"step": 2460
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.9968672628972766,
|
| 1750 |
+
"grad_norm": 1.5404280086355402,
|
| 1751 |
+
"learning_rate": 5.300485404239999e-06,
|
| 1752 |
+
"loss": 0.9628,
|
| 1753 |
+
"step": 2470
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.999292607750998,
|
| 1757 |
+
"eval_loss": 0.8751075863838196,
|
| 1758 |
+
"eval_runtime": 481.67,
|
| 1759 |
+
"eval_samples_per_second": 25.254,
|
| 1760 |
+
"eval_steps_per_second": 12.627,
|
| 1761 |
+
"step": 2473
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 2.0049517457430146,
|
| 1765 |
+
"grad_norm": 1.8577507088673693,
|
| 1766 |
+
"learning_rate": 5.223550612074497e-06,
|
| 1767 |
+
"loss": 0.8752,
|
| 1768 |
+
"step": 2480
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.0130362285887524,
|
| 1772 |
+
"grad_norm": 1.5570324756102374,
|
| 1773 |
+
"learning_rate": 5.146980374689192e-06,
|
| 1774 |
+
"loss": 0.8398,
|
| 1775 |
+
"step": 2490
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 2.0211207114344902,
|
| 1779 |
+
"grad_norm": 1.645225536576169,
|
| 1780 |
+
"learning_rate": 5.070780536183698e-06,
|
| 1781 |
+
"loss": 0.856,
|
| 1782 |
+
"step": 2500
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 2.0292051942802285,
|
| 1786 |
+
"grad_norm": 1.6698633554870226,
|
| 1787 |
+
"learning_rate": 4.99495691238755e-06,
|
| 1788 |
+
"loss": 0.8365,
|
| 1789 |
+
"step": 2510
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 2.0372896771259663,
|
| 1793 |
+
"grad_norm": 2.010967933907663,
|
| 1794 |
+
"learning_rate": 4.9195152904162865e-06,
|
| 1795 |
+
"loss": 0.8308,
|
| 1796 |
+
"step": 2520
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 2.045374159971704,
|
| 1800 |
+
"grad_norm": 1.4592026658551123,
|
| 1801 |
+
"learning_rate": 4.844461428229782e-06,
|
| 1802 |
+
"loss": 0.8387,
|
| 1803 |
+
"step": 2530
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 2.0534586428174424,
|
| 1807 |
+
"grad_norm": 1.9716723547932462,
|
| 1808 |
+
"learning_rate": 4.769801054192776e-06,
|
| 1809 |
+
"loss": 0.8374,
|
| 1810 |
+
"step": 2540
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 2.06154312566318,
|
| 1814 |
+
"grad_norm": 1.6334367414667887,
|
| 1815 |
+
"learning_rate": 4.695539866637653e-06,
|
| 1816 |
+
"loss": 0.8587,
|
| 1817 |
+
"step": 2550
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 2.069627608508918,
|
| 1821 |
+
"grad_norm": 1.713926689166813,
|
| 1822 |
+
"learning_rate": 4.6216835334295385e-06,
|
| 1823 |
+
"loss": 0.8376,
|
| 1824 |
+
"step": 2560
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 2.0777120913546563,
|
| 1828 |
+
"grad_norm": 1.5714175555320091,
|
| 1829 |
+
"learning_rate": 4.548237691533699e-06,
|
| 1830 |
+
"loss": 0.8346,
|
| 1831 |
+
"step": 2570
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 2.085796574200394,
|
| 1835 |
+
"grad_norm": 1.4811489223457255,
|
| 1836 |
+
"learning_rate": 4.475207946585328e-06,
|
| 1837 |
+
"loss": 0.8473,
|
| 1838 |
+
"step": 2580
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 2.093881057046132,
|
| 1842 |
+
"grad_norm": 1.4400201402098334,
|
| 1843 |
+
"learning_rate": 4.402599872461678e-06,
|
| 1844 |
+
"loss": 0.8309,
|
| 1845 |
+
"step": 2590
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 2.10196553989187,
|
| 1849 |
+
"grad_norm": 1.5527150219002093,
|
| 1850 |
+
"learning_rate": 4.330419010856661e-06,
|
| 1851 |
+
"loss": 0.8312,
|
| 1852 |
+
"step": 2600
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 2.110050022737608,
|
| 1856 |
+
"grad_norm": 1.4540137626455856,
|
| 1857 |
+
"learning_rate": 4.258670870857894e-06,
|
| 1858 |
+
"loss": 0.8461,
|
| 1859 |
+
"step": 2610
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 2.118134505583346,
|
| 1863 |
+
"grad_norm": 1.5200526871374724,
|
| 1864 |
+
"learning_rate": 4.187360928526198e-06,
|
| 1865 |
+
"loss": 0.8353,
|
| 1866 |
+
"step": 2620
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 2.126218988429084,
|
| 1870 |
+
"grad_norm": 1.487656190760893,
|
| 1871 |
+
"learning_rate": 4.116494626477684e-06,
|
| 1872 |
+
"loss": 0.842,
|
| 1873 |
+
"step": 2630
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 2.134303471274822,
|
| 1877 |
+
"grad_norm": 1.4541876796717628,
|
| 1878 |
+
"learning_rate": 4.046077373468325e-06,
|
| 1879 |
+
"loss": 0.8285,
|
| 1880 |
+
"step": 2640
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 2.1423879541205597,
|
| 1884 |
+
"grad_norm": 1.515080712913025,
|
| 1885 |
+
"learning_rate": 3.976114543981148e-06,
|
| 1886 |
+
"loss": 0.8278,
|
| 1887 |
+
"step": 2650
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 2.150472436966298,
|
| 1891 |
+
"grad_norm": 1.5925627792233104,
|
| 1892 |
+
"learning_rate": 3.906611477816054e-06,
|
| 1893 |
+
"loss": 0.8382,
|
| 1894 |
+
"step": 2660
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 2.158556919812036,
|
| 1898 |
+
"grad_norm": 1.4749306746231339,
|
| 1899 |
+
"learning_rate": 3.837573479682236e-06,
|
| 1900 |
+
"loss": 0.8453,
|
| 1901 |
+
"step": 2670
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 2.1666414026577736,
|
| 1905 |
+
"grad_norm": 1.888042329530717,
|
| 1906 |
+
"learning_rate": 3.769005818793329e-06,
|
| 1907 |
+
"loss": 0.854,
|
| 1908 |
+
"step": 2680
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 2.174725885503512,
|
| 1912 |
+
"grad_norm": 1.598037794600047,
|
| 1913 |
+
"learning_rate": 3.7009137284652386e-06,
|
| 1914 |
+
"loss": 0.8519,
|
| 1915 |
+
"step": 2690
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 2.1828103683492497,
|
| 1919 |
+
"grad_norm": 1.5540837615094885,
|
| 1920 |
+
"learning_rate": 3.633302405716712e-06,
|
| 1921 |
+
"loss": 0.8397,
|
| 1922 |
+
"step": 2700
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.1908948511949875,
|
| 1926 |
+
"grad_norm": 1.430485289060877,
|
| 1927 |
+
"learning_rate": 3.5661770108726914e-06,
|
| 1928 |
+
"loss": 0.8271,
|
| 1929 |
+
"step": 2710
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 2.1989793340407258,
|
| 1933 |
+
"grad_norm": 2.401835949374892,
|
| 1934 |
+
"learning_rate": 3.4995426671704493e-06,
|
| 1935 |
+
"loss": 0.8335,
|
| 1936 |
+
"step": 2720
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 2.2070638168864636,
|
| 1940 |
+
"grad_norm": 1.506353292247366,
|
| 1941 |
+
"learning_rate": 3.433404460368587e-06,
|
| 1942 |
+
"loss": 0.828,
|
| 1943 |
+
"step": 2730
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 2.2151482997322014,
|
| 1947 |
+
"grad_norm": 1.4406717845115946,
|
| 1948 |
+
"learning_rate": 3.3677674383588476e-06,
|
| 1949 |
+
"loss": 0.8315,
|
| 1950 |
+
"step": 2740
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 2.2232327825779397,
|
| 1954 |
+
"grad_norm": 1.5393945850323205,
|
| 1955 |
+
"learning_rate": 3.302636610780855e-06,
|
| 1956 |
+
"loss": 0.8504,
|
| 1957 |
+
"step": 2750
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 2.2313172654236775,
|
| 1961 |
+
"grad_norm": 1.7257558230682333,
|
| 1962 |
+
"learning_rate": 3.238016948639772e-06,
|
| 1963 |
+
"loss": 0.8232,
|
| 1964 |
+
"step": 2760
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.2394017482694153,
|
| 1968 |
+
"grad_norm": 1.8326756661400847,
|
| 1969 |
+
"learning_rate": 3.1739133839268698e-06,
|
| 1970 |
+
"loss": 0.8154,
|
| 1971 |
+
"step": 2770
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 2.2474862311151536,
|
| 1975 |
+
"grad_norm": 1.5269518503128512,
|
| 1976 |
+
"learning_rate": 3.110330809243134e-06,
|
| 1977 |
+
"loss": 0.8317,
|
| 1978 |
+
"step": 2780
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.2555707139608914,
|
| 1982 |
+
"grad_norm": 1.504166909878008,
|
| 1983 |
+
"learning_rate": 3.0472740774258157e-06,
|
| 1984 |
+
"loss": 0.8368,
|
| 1985 |
+
"step": 2790
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 2.263655196806629,
|
| 1989 |
+
"grad_norm": 1.480047137104623,
|
| 1990 |
+
"learning_rate": 2.9847480011780607e-06,
|
| 1991 |
+
"loss": 0.8409,
|
| 1992 |
+
"step": 2800
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 2.2717396796523674,
|
| 1996 |
+
"grad_norm": 1.492023552078346,
|
| 1997 |
+
"learning_rate": 2.922757352701595e-06,
|
| 1998 |
+
"loss": 0.8243,
|
| 1999 |
+
"step": 2810
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 2.2798241624981053,
|
| 2003 |
+
"grad_norm": 1.467055149697424,
|
| 2004 |
+
"learning_rate": 2.861306863332475e-06,
|
| 2005 |
+
"loss": 0.8289,
|
| 2006 |
+
"step": 2820
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.287908645343843,
|
| 2010 |
+
"grad_norm": 1.504514345406056,
|
| 2011 |
+
"learning_rate": 2.8004012231799905e-06,
|
| 2012 |
+
"loss": 0.8375,
|
| 2013 |
+
"step": 2830
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.295993128189581,
|
| 2017 |
+
"grad_norm": 1.5091792435489357,
|
| 2018 |
+
"learning_rate": 2.740045080768694e-06,
|
| 2019 |
+
"loss": 0.8233,
|
| 2020 |
+
"step": 2840
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 2.304077611035319,
|
| 2024 |
+
"grad_norm": 1.4619080284602382,
|
| 2025 |
+
"learning_rate": 2.6802430426836113e-06,
|
| 2026 |
+
"loss": 0.8356,
|
| 2027 |
+
"step": 2850
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.312162093881057,
|
| 2031 |
+
"grad_norm": 1.4085751552174153,
|
| 2032 |
+
"learning_rate": 2.620999673218656e-06,
|
| 2033 |
+
"loss": 0.8156,
|
| 2034 |
+
"step": 2860
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 2.3202465767267952,
|
| 2038 |
+
"grad_norm": 1.4755258769825808,
|
| 2039 |
+
"learning_rate": 2.5623194940282526e-06,
|
| 2040 |
+
"loss": 0.8353,
|
| 2041 |
+
"step": 2870
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 2.328331059572533,
|
| 2045 |
+
"grad_norm": 1.5852343601430656,
|
| 2046 |
+
"learning_rate": 2.504206983782248e-06,
|
| 2047 |
+
"loss": 0.8133,
|
| 2048 |
+
"step": 2880
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.336415542418271,
|
| 2052 |
+
"grad_norm": 1.4903107631764194,
|
| 2053 |
+
"learning_rate": 2.446666577824068e-06,
|
| 2054 |
+
"loss": 0.8459,
|
| 2055 |
+
"step": 2890
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.3445000252640087,
|
| 2059 |
+
"grad_norm": 1.523719484539125,
|
| 2060 |
+
"learning_rate": 2.389702667832202e-06,
|
| 2061 |
+
"loss": 0.8285,
|
| 2062 |
+
"step": 2900
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 2.352584508109747,
|
| 2066 |
+
"grad_norm": 1.457321496284554,
|
| 2067 |
+
"learning_rate": 2.3333196014850246e-06,
|
| 2068 |
+
"loss": 0.8304,
|
| 2069 |
+
"step": 2910
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.3606689909554848,
|
| 2073 |
+
"grad_norm": 1.537434676857527,
|
| 2074 |
+
"learning_rate": 2.277521682128947e-06,
|
| 2075 |
+
"loss": 0.829,
|
| 2076 |
+
"step": 2920
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 2.3687534738012226,
|
| 2080 |
+
"grad_norm": 1.4707817420987006,
|
| 2081 |
+
"learning_rate": 2.2223131684499932e-06,
|
| 2082 |
+
"loss": 0.8372,
|
| 2083 |
+
"step": 2930
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.376837956646961,
|
| 2087 |
+
"grad_norm": 1.46749047915079,
|
| 2088 |
+
"learning_rate": 2.1676982741487427e-06,
|
| 2089 |
+
"loss": 0.8222,
|
| 2090 |
+
"step": 2940
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.3849224394926987,
|
| 2094 |
+
"grad_norm": 1.518122852634397,
|
| 2095 |
+
"learning_rate": 2.113681167618736e-06,
|
| 2096 |
+
"loss": 0.8401,
|
| 2097 |
+
"step": 2950
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.3930069223384365,
|
| 2101 |
+
"grad_norm": 1.8575848589445734,
|
| 2102 |
+
"learning_rate": 2.060265971628338e-06,
|
| 2103 |
+
"loss": 0.8339,
|
| 2104 |
+
"step": 2960
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 2.4010914051841747,
|
| 2108 |
+
"grad_norm": 1.5601145654381285,
|
| 2109 |
+
"learning_rate": 2.0074567630060514e-06,
|
| 2110 |
+
"loss": 0.8154,
|
| 2111 |
+
"step": 2970
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.4091758880299126,
|
| 2115 |
+
"grad_norm": 1.530898387002521,
|
| 2116 |
+
"learning_rate": 1.955257572329379e-06,
|
| 2117 |
+
"loss": 0.823,
|
| 2118 |
+
"step": 2980
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 2.4172603708756504,
|
| 2122 |
+
"grad_norm": 1.6224545445427798,
|
| 2123 |
+
"learning_rate": 1.9036723836171899e-06,
|
| 2124 |
+
"loss": 0.8145,
|
| 2125 |
+
"step": 2990
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 2.4253448537213886,
|
| 2129 |
+
"grad_norm": 1.4013679708594033,
|
| 2130 |
+
"learning_rate": 1.8527051340256397e-06,
|
| 2131 |
+
"loss": 0.8215,
|
| 2132 |
+
"step": 3000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.4334293365671265,
|
| 2136 |
+
"grad_norm": 1.5692785609667004,
|
| 2137 |
+
"learning_rate": 1.8023597135476923e-06,
|
| 2138 |
+
"loss": 0.8241,
|
| 2139 |
+
"step": 3010
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 2.4415138194128643,
|
| 2143 |
+
"grad_norm": 1.5126974695662643,
|
| 2144 |
+
"learning_rate": 1.752639964716193e-06,
|
| 2145 |
+
"loss": 0.8421,
|
| 2146 |
+
"step": 3020
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 2.4495983022586025,
|
| 2150 |
+
"grad_norm": 1.6242742569822604,
|
| 2151 |
+
"learning_rate": 1.7035496823106247e-06,
|
| 2152 |
+
"loss": 0.8141,
|
| 2153 |
+
"step": 3030
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 2.4576827851043404,
|
| 2157 |
+
"grad_norm": 1.4628790110692993,
|
| 2158 |
+
"learning_rate": 1.6550926130674527e-06,
|
| 2159 |
+
"loss": 0.8184,
|
| 2160 |
+
"step": 3040
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 2.465767267950078,
|
| 2164 |
+
"grad_norm": 1.4807837431822446,
|
| 2165 |
+
"learning_rate": 1.607272455394172e-06,
|
| 2166 |
+
"loss": 0.8202,
|
| 2167 |
+
"step": 3050
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 2.4738517507958164,
|
| 2171 |
+
"grad_norm": 1.5539937903441552,
|
| 2172 |
+
"learning_rate": 1.5600928590870402e-06,
|
| 2173 |
+
"loss": 0.8391,
|
| 2174 |
+
"step": 3060
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 2.4819362336415542,
|
| 2178 |
+
"grad_norm": 1.6677495360703212,
|
| 2179 |
+
"learning_rate": 1.5135574250524898e-06,
|
| 2180 |
+
"loss": 0.8436,
|
| 2181 |
+
"step": 3070
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.490020716487292,
|
| 2185 |
+
"grad_norm": 1.53769857798961,
|
| 2186 |
+
"learning_rate": 1.467669705032323e-06,
|
| 2187 |
+
"loss": 0.8263,
|
| 2188 |
+
"step": 3080
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 2.4981051993330303,
|
| 2192 |
+
"grad_norm": 1.4732928239069325,
|
| 2193 |
+
"learning_rate": 1.422433201332607e-06,
|
| 2194 |
+
"loss": 0.8284,
|
| 2195 |
+
"step": 3090
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 2.506189682178768,
|
| 2199 |
+
"grad_norm": 1.5928757648188723,
|
| 2200 |
+
"learning_rate": 1.3778513665563786e-06,
|
| 2201 |
+
"loss": 0.8319,
|
| 2202 |
+
"step": 3100
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 2.514274165024506,
|
| 2206 |
+
"grad_norm": 1.4230928346180836,
|
| 2207 |
+
"learning_rate": 1.3339276033401283e-06,
|
| 2208 |
+
"loss": 0.8052,
|
| 2209 |
+
"step": 3110
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 2.522358647870244,
|
| 2213 |
+
"grad_norm": 1.4772661299744003,
|
| 2214 |
+
"learning_rate": 1.290665264094093e-06,
|
| 2215 |
+
"loss": 0.8241,
|
| 2216 |
+
"step": 3120
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 2.530443130715982,
|
| 2220 |
+
"grad_norm": 1.522091825661006,
|
| 2221 |
+
"learning_rate": 1.2480676507464051e-06,
|
| 2222 |
+
"loss": 0.8106,
|
| 2223 |
+
"step": 3130
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 2.53852761356172,
|
| 2227 |
+
"grad_norm": 1.525599170654266,
|
| 2228 |
+
"learning_rate": 1.2061380144910572e-06,
|
| 2229 |
+
"loss": 0.8166,
|
| 2230 |
+
"step": 3140
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 2.5466120964074577,
|
| 2234 |
+
"grad_norm": 1.4929327017491605,
|
| 2235 |
+
"learning_rate": 1.1648795555397719e-06,
|
| 2236 |
+
"loss": 0.8251,
|
| 2237 |
+
"step": 3150
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 2.554696579253196,
|
| 2241 |
+
"grad_norm": 1.5920001415947864,
|
| 2242 |
+
"learning_rate": 1.1242954228777513e-06,
|
| 2243 |
+
"loss": 0.8268,
|
| 2244 |
+
"step": 3160
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 2.5627810620989337,
|
| 2248 |
+
"grad_norm": 1.5252651359986042,
|
| 2249 |
+
"learning_rate": 1.08438871402333e-06,
|
| 2250 |
+
"loss": 0.831,
|
| 2251 |
+
"step": 3170
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 2.570865544944672,
|
| 2255 |
+
"grad_norm": 1.6461347768103347,
|
| 2256 |
+
"learning_rate": 1.04516247479157e-06,
|
| 2257 |
+
"loss": 0.8239,
|
| 2258 |
+
"step": 3180
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 2.57895002779041,
|
| 2262 |
+
"grad_norm": 1.490863354097273,
|
| 2263 |
+
"learning_rate": 1.006619699061785e-06,
|
| 2264 |
+
"loss": 0.823,
|
| 2265 |
+
"step": 3190
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 2.5870345106361476,
|
| 2269 |
+
"grad_norm": 1.5158841203253022,
|
| 2270 |
+
"learning_rate": 9.687633285490395e-07,
|
| 2271 |
+
"loss": 0.8333,
|
| 2272 |
+
"step": 3200
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 2.5951189934818855,
|
| 2276 |
+
"grad_norm": 1.4861408651974157,
|
| 2277 |
+
"learning_rate": 9.315962525796374e-07,
|
| 2278 |
+
"loss": 0.8178,
|
| 2279 |
+
"step": 3210
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 2.6032034763276237,
|
| 2283 |
+
"grad_norm": 1.4847726389856295,
|
| 2284 |
+
"learning_rate": 8.951213078705811e-07,
|
| 2285 |
+
"loss": 0.8244,
|
| 2286 |
+
"step": 3220
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 2.6112879591733615,
|
| 2290 |
+
"grad_norm": 1.4579228976188288,
|
| 2291 |
+
"learning_rate": 8.593412783130805e-07,
|
| 2292 |
+
"loss": 0.8116,
|
| 2293 |
+
"step": 3230
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 2.6193724420191,
|
| 2297 |
+
"grad_norm": 1.4309284818257009,
|
| 2298 |
+
"learning_rate": 8.24258894760066e-07,
|
| 2299 |
+
"loss": 0.8233,
|
| 2300 |
+
"step": 3240
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 2.6274569248648376,
|
| 2304 |
+
"grad_norm": 1.481662266621092,
|
| 2305 |
+
"learning_rate": 7.898768348177643e-07,
|
| 2306 |
+
"loss": 0.8393,
|
| 2307 |
+
"step": 3250
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 2.6355414077105754,
|
| 2311 |
+
"grad_norm": 1.42582017885812,
|
| 2312 |
+
"learning_rate": 7.561977226413341e-07,
|
| 2313 |
+
"loss": 0.8344,
|
| 2314 |
+
"step": 3260
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 2.6436258905563133,
|
| 2318 |
+
"grad_norm": 1.4203791210214531,
|
| 2319 |
+
"learning_rate": 7.23224128734582e-07,
|
| 2320 |
+
"loss": 0.821,
|
| 2321 |
+
"step": 3270
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.6517103734020515,
|
| 2325 |
+
"grad_norm": 1.4780417621137758,
|
| 2326 |
+
"learning_rate": 6.909585697537758e-07,
|
| 2327 |
+
"loss": 0.8353,
|
| 2328 |
+
"step": 3280
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 2.6597948562477893,
|
| 2332 |
+
"grad_norm": 1.4466612391449976,
|
| 2333 |
+
"learning_rate": 6.594035083155581e-07,
|
| 2334 |
+
"loss": 0.8268,
|
| 2335 |
+
"step": 3290
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 2.6678793390935276,
|
| 2339 |
+
"grad_norm": 1.4584592752103582,
|
| 2340 |
+
"learning_rate": 6.285613528089962e-07,
|
| 2341 |
+
"loss": 0.8164,
|
| 2342 |
+
"step": 3300
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 2.6759638219392654,
|
| 2346 |
+
"grad_norm": 1.487514724946772,
|
| 2347 |
+
"learning_rate": 5.98434457211765e-07,
|
| 2348 |
+
"loss": 0.8027,
|
| 2349 |
+
"step": 3310
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 2.6840483047850032,
|
| 2353 |
+
"grad_norm": 1.4294666752405771,
|
| 2354 |
+
"learning_rate": 5.690251209104802e-07,
|
| 2355 |
+
"loss": 0.8105,
|
| 2356 |
+
"step": 3320
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 2.692132787630741,
|
| 2360 |
+
"grad_norm": 1.4638925402226952,
|
| 2361 |
+
"learning_rate": 5.403355885252104e-07,
|
| 2362 |
+
"loss": 0.8135,
|
| 2363 |
+
"step": 3330
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 2.7002172704764793,
|
| 2367 |
+
"grad_norm": 1.4458763488108235,
|
| 2368 |
+
"learning_rate": 5.123680497381444e-07,
|
| 2369 |
+
"loss": 0.8102,
|
| 2370 |
+
"step": 3340
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 2.708301753322217,
|
| 2374 |
+
"grad_norm": 1.4903596037049076,
|
| 2375 |
+
"learning_rate": 4.851246391264819e-07,
|
| 2376 |
+
"loss": 0.8152,
|
| 2377 |
+
"step": 3350
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 2.7163862361679554,
|
| 2381 |
+
"grad_norm": 1.4429528216246368,
|
| 2382 |
+
"learning_rate": 4.5860743599951186e-07,
|
| 2383 |
+
"loss": 0.8121,
|
| 2384 |
+
"step": 3360
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 2.724470719013693,
|
| 2388 |
+
"grad_norm": 1.452035259914063,
|
| 2389 |
+
"learning_rate": 4.328184642399036e-07,
|
| 2390 |
+
"loss": 0.821,
|
| 2391 |
+
"step": 3370
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.732555201859431,
|
| 2395 |
+
"grad_norm": 1.5303877229228735,
|
| 2396 |
+
"learning_rate": 4.077596921492533e-07,
|
| 2397 |
+
"loss": 0.8145,
|
| 2398 |
+
"step": 3380
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 2.740639684705169,
|
| 2402 |
+
"grad_norm": 1.4449405328561624,
|
| 2403 |
+
"learning_rate": 3.834330322978397e-07,
|
| 2404 |
+
"loss": 0.8214,
|
| 2405 |
+
"step": 3390
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 2.748724167550907,
|
| 2409 |
+
"grad_norm": 1.4371584227135465,
|
| 2410 |
+
"learning_rate": 3.598403413786611e-07,
|
| 2411 |
+
"loss": 0.8131,
|
| 2412 |
+
"step": 3400
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 2.756808650396645,
|
| 2416 |
+
"grad_norm": 1.4632980675092546,
|
| 2417 |
+
"learning_rate": 3.3698342006572294e-07,
|
| 2418 |
+
"loss": 0.8244,
|
| 2419 |
+
"step": 3410
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 2.764893133242383,
|
| 2423 |
+
"grad_norm": 1.4500755832832954,
|
| 2424 |
+
"learning_rate": 3.148640128766056e-07,
|
| 2425 |
+
"loss": 0.823,
|
| 2426 |
+
"step": 3420
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 2.772977616088121,
|
| 2430 |
+
"grad_norm": 1.4751477866660623,
|
| 2431 |
+
"learning_rate": 2.934838080393154e-07,
|
| 2432 |
+
"loss": 0.8211,
|
| 2433 |
+
"step": 3430
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 2.781062098933859,
|
| 2437 |
+
"grad_norm": 1.4653755137740456,
|
| 2438 |
+
"learning_rate": 2.7284443736343203e-07,
|
| 2439 |
+
"loss": 0.8024,
|
| 2440 |
+
"step": 3440
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 2.7891465817795966,
|
| 2444 |
+
"grad_norm": 1.4089563044736344,
|
| 2445 |
+
"learning_rate": 2.52947476115567e-07,
|
| 2446 |
+
"loss": 0.8228,
|
| 2447 |
+
"step": 3450
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 2.797231064625335,
|
| 2451 |
+
"grad_norm": 1.460696621649454,
|
| 2452 |
+
"learning_rate": 2.3379444289913344e-07,
|
| 2453 |
+
"loss": 0.8184,
|
| 2454 |
+
"step": 3460
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 2.8053155474710727,
|
| 2458 |
+
"grad_norm": 1.4693334824298931,
|
| 2459 |
+
"learning_rate": 2.153867995384351e-07,
|
| 2460 |
+
"loss": 0.8224,
|
| 2461 |
+
"step": 3470
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 2.8134000303168105,
|
| 2465 |
+
"grad_norm": 1.4469954005038157,
|
| 2466 |
+
"learning_rate": 1.9772595096710477e-07,
|
| 2467 |
+
"loss": 0.8373,
|
| 2468 |
+
"step": 3480
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 2.821484513162549,
|
| 2472 |
+
"grad_norm": 1.4331150676229163,
|
| 2473 |
+
"learning_rate": 1.8081324512086663e-07,
|
| 2474 |
+
"loss": 0.8185,
|
| 2475 |
+
"step": 3490
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 2.8295689960082866,
|
| 2479 |
+
"grad_norm": 1.5335384382024873,
|
| 2480 |
+
"learning_rate": 1.6464997283466067e-07,
|
| 2481 |
+
"loss": 0.8124,
|
| 2482 |
+
"step": 3500
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 2.8376534788540244,
|
| 2486 |
+
"grad_norm": 1.4445147972537609,
|
| 2487 |
+
"learning_rate": 1.492373677441228e-07,
|
| 2488 |
+
"loss": 0.8145,
|
| 2489 |
+
"step": 3510
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 2.8457379616997627,
|
| 2493 |
+
"grad_norm": 1.4976188260457166,
|
| 2494 |
+
"learning_rate": 1.3457660619142887e-07,
|
| 2495 |
+
"loss": 0.8163,
|
| 2496 |
+
"step": 3520
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 2.8538224445455005,
|
| 2500 |
+
"grad_norm": 1.439452743377751,
|
| 2501 |
+
"learning_rate": 1.2066880713550888e-07,
|
| 2502 |
+
"loss": 0.829,
|
| 2503 |
+
"step": 3530
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 2.8619069273912383,
|
| 2507 |
+
"grad_norm": 1.524984754735583,
|
| 2508 |
+
"learning_rate": 1.0751503206665071e-07,
|
| 2509 |
+
"loss": 0.8236,
|
| 2510 |
+
"step": 3540
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 2.8699914102369766,
|
| 2514 |
+
"grad_norm": 1.448229914768272,
|
| 2515 |
+
"learning_rate": 9.511628492547609e-08,
|
| 2516 |
+
"loss": 0.8223,
|
| 2517 |
+
"step": 3550
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 2.8780758930827144,
|
| 2521 |
+
"grad_norm": 1.4915344957228824,
|
| 2522 |
+
"learning_rate": 8.347351202632525e-08,
|
| 2523 |
+
"loss": 0.843,
|
| 2524 |
+
"step": 3560
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 2.886160375928452,
|
| 2528 |
+
"grad_norm": 1.4891660841319714,
|
| 2529 |
+
"learning_rate": 7.258760198502246e-08,
|
| 2530 |
+
"loss": 0.8173,
|
| 2531 |
+
"step": 3570
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 2.89424485877419,
|
| 2535 |
+
"grad_norm": 1.4485487573496472,
|
| 2536 |
+
"learning_rate": 6.245938565105803e-08,
|
| 2537 |
+
"loss": 0.8299,
|
| 2538 |
+
"step": 3580
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 2.9023293416199283,
|
| 2542 |
+
"grad_norm": 1.452602418516034,
|
| 2543 |
+
"learning_rate": 5.308963604417572e-08,
|
| 2544 |
+
"loss": 0.8216,
|
| 2545 |
+
"step": 3590
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 2.910413824465666,
|
| 2549 |
+
"grad_norm": 1.4554407329371093,
|
| 2550 |
+
"learning_rate": 4.447906829537219e-08,
|
| 2551 |
+
"loss": 0.8284,
|
| 2552 |
+
"step": 3600
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 2.9184983073114044,
|
| 2556 |
+
"grad_norm": 1.4918607001029844,
|
| 2557 |
+
"learning_rate": 3.6628339592313935e-08,
|
| 2558 |
+
"loss": 0.8012,
|
| 2559 |
+
"step": 3610
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 2.926582790157142,
|
| 2563 |
+
"grad_norm": 1.4229324193215207,
|
| 2564 |
+
"learning_rate": 2.95380491291819e-08,
|
| 2565 |
+
"loss": 0.8401,
|
| 2566 |
+
"step": 3620
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 2.93466727300288,
|
| 2570 |
+
"grad_norm": 1.4288366788035922,
|
| 2571 |
+
"learning_rate": 2.320873806093804e-08,
|
| 2572 |
+
"loss": 0.8228,
|
| 2573 |
+
"step": 3630
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 2.942751755848618,
|
| 2577 |
+
"grad_norm": 1.4724134547959333,
|
| 2578 |
+
"learning_rate": 1.764088946201947e-08,
|
| 2579 |
+
"loss": 0.8064,
|
| 2580 |
+
"step": 3640
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 2.950836238694356,
|
| 2584 |
+
"grad_norm": 1.4984479737935563,
|
| 2585 |
+
"learning_rate": 1.2834928289472415e-08,
|
| 2586 |
+
"loss": 0.81,
|
| 2587 |
+
"step": 3650
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 2.958920721540094,
|
| 2591 |
+
"grad_norm": 1.4666816312445612,
|
| 2592 |
+
"learning_rate": 8.79122135051591e-09,
|
| 2593 |
+
"loss": 0.822,
|
| 2594 |
+
"step": 3660
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 2.967005204385832,
|
| 2598 |
+
"grad_norm": 1.445201429621803,
|
| 2599 |
+
"learning_rate": 5.510077274547554e-09,
|
| 2600 |
+
"loss": 0.8271,
|
| 2601 |
+
"step": 3670
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 2.97508968723157,
|
| 2605 |
+
"grad_norm": 1.4460059967392547,
|
| 2606 |
+
"learning_rate": 2.9917464895856673e-09,
|
| 2607 |
+
"loss": 0.8389,
|
| 2608 |
+
"step": 3680
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 2.983174170077308,
|
| 2612 |
+
"grad_norm": 1.435390156627942,
|
| 2613 |
+
"learning_rate": 1.2364212031579226e-09,
|
| 2614 |
+
"loss": 0.8294,
|
| 2615 |
+
"step": 3690
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 2.9912586529230456,
|
| 2619 |
+
"grad_norm": 1.5066669703721747,
|
| 2620 |
+
"learning_rate": 2.442353876297432e-10,
|
| 2621 |
+
"loss": 0.801,
|
| 2622 |
+
"step": 3700
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 2.997726239199636,
|
| 2626 |
+
"eval_loss": 0.8224219083786011,
|
| 2627 |
+
"eval_runtime": 474.463,
|
| 2628 |
+
"eval_samples_per_second": 25.637,
|
| 2629 |
+
"eval_steps_per_second": 12.819,
|
| 2630 |
+
"step": 3708
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 2.997726239199636,
|
| 2634 |
+
"step": 3708,
|
| 2635 |
+
"total_flos": 0.0,
|
| 2636 |
+
"train_loss": 1.0273753281164324,
|
| 2637 |
+
"train_runtime": 58675.1239,
|
| 2638 |
+
"train_samples_per_second": 8.095,
|
| 2639 |
+
"train_steps_per_second": 0.063
|
| 2640 |
+
}
|
| 2641 |
+
],
|
| 2642 |
+
"logging_steps": 10,
|
| 2643 |
+
"max_steps": 3708,
|
| 2644 |
+
"num_input_tokens_seen": 0,
|
| 2645 |
+
"num_train_epochs": 3,
|
| 2646 |
+
"save_steps": 100,
|
| 2647 |
+
"total_flos": 0.0,
|
| 2648 |
+
"train_batch_size": 8,
|
| 2649 |
+
"trial_name": null,
|
| 2650 |
+
"trial_params": null
|
| 2651 |
+
}
|