Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: gemma
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: google/gemma-7b
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: gemma7b-summarize-gpt4o-32k
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# gemma7b-summarize-gpt4o-32k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 4.7170
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0002
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
+
- eval_batch_size: 4
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 2
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 16
|
| 50 |
+
- total_eval_batch_size: 8
|
| 51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
+
- lr_scheduler_type: cosine
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 54 |
+
- num_epochs: 10
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
+
| 1.0031 | 0.9975 | 203 | 2.5299 |
|
| 61 |
+
| 0.8685 | 2.0 | 407 | 2.4260 |
|
| 62 |
+
| 0.8 | 2.9975 | 610 | 2.5051 |
|
| 63 |
+
| 0.6938 | 4.0 | 814 | 2.6558 |
|
| 64 |
+
| 0.5865 | 4.9975 | 1017 | 2.9163 |
|
| 65 |
+
| 0.4821 | 6.0 | 1221 | 3.3285 |
|
| 66 |
+
| 0.3899 | 6.9975 | 1424 | 3.8984 |
|
| 67 |
+
| 0.3321 | 8.0 | 1628 | 4.4348 |
|
| 68 |
+
| 0.3089 | 8.9975 | 1831 | 4.6891 |
|
| 69 |
+
| 0.3016 | 9.9754 | 2030 | 4.7170 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- PEFT 0.10.0
|
| 75 |
+
- Transformers 4.40.0
|
| 76 |
+
- Pytorch 2.2.1+cu121
|
| 77 |
+
- Datasets 2.18.0
|
| 78 |
+
- Tokenizers 0.19.1
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 50056096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:125f331ca9a8684aec05b7fd7ab37fb0f3ec43d64414a7c4aa9dfc598b180031
|
| 3 |
size 50056096
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.975429975429975,
|
| 3 |
+
"total_flos": 1.5518062706111283e+18,
|
| 4 |
+
"train_loss": 1.3545548074938394,
|
| 5 |
+
"train_runtime": 13205.4864,
|
| 6 |
+
"train_samples": 29997,
|
| 7 |
+
"train_samples_per_second": 2.465,
|
| 8 |
+
"train_steps_per_second": 0.154
|
| 9 |
+
}
|
runs/Jun08_17-03-13_user-WS-C621E-SAGE-Series/events.out.tfevents.1717833873.user-WS-C621E-SAGE-Series.13572.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5379c8320c1ec38d3a342189d2539ae48a02598f593cdad40d5c6748b67dbf2
|
| 3 |
+
size 94269
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.975429975429975,
|
| 3 |
+
"total_flos": 1.5518062706111283e+18,
|
| 4 |
+
"train_loss": 1.3545548074938394,
|
| 5 |
+
"train_runtime": 13205.4864,
|
| 6 |
+
"train_samples": 29997,
|
| 7 |
+
"train_samples_per_second": 2.465,
|
| 8 |
+
"train_steps_per_second": 0.154
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2959 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.975429975429975,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2030,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004914004914004914,
|
| 13 |
+
"grad_norm": 552.0,
|
| 14 |
+
"learning_rate": 9.852216748768474e-07,
|
| 15 |
+
"loss": 51.7129,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02457002457002457,
|
| 20 |
+
"grad_norm": 404.0,
|
| 21 |
+
"learning_rate": 4.926108374384237e-06,
|
| 22 |
+
"loss": 45.7069,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04914004914004914,
|
| 27 |
+
"grad_norm": 388.0,
|
| 28 |
+
"learning_rate": 9.852216748768475e-06,
|
| 29 |
+
"loss": 46.9702,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07371007371007371,
|
| 34 |
+
"grad_norm": 292.0,
|
| 35 |
+
"learning_rate": 1.4778325123152711e-05,
|
| 36 |
+
"loss": 37.466,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09828009828009827,
|
| 41 |
+
"grad_norm": 79.5,
|
| 42 |
+
"learning_rate": 1.970443349753695e-05,
|
| 43 |
+
"loss": 28.4521,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12285012285012285,
|
| 48 |
+
"grad_norm": 47.25,
|
| 49 |
+
"learning_rate": 2.4630541871921184e-05,
|
| 50 |
+
"loss": 23.8617,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.14742014742014742,
|
| 55 |
+
"grad_norm": 26.625,
|
| 56 |
+
"learning_rate": 2.9556650246305422e-05,
|
| 57 |
+
"loss": 22.4994,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.171990171990172,
|
| 62 |
+
"grad_norm": 12.625,
|
| 63 |
+
"learning_rate": 3.4482758620689657e-05,
|
| 64 |
+
"loss": 20.8855,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.19656019656019655,
|
| 69 |
+
"grad_norm": 7.84375,
|
| 70 |
+
"learning_rate": 3.94088669950739e-05,
|
| 71 |
+
"loss": 19.2048,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.22113022113022113,
|
| 76 |
+
"grad_norm": 8.5,
|
| 77 |
+
"learning_rate": 4.433497536945813e-05,
|
| 78 |
+
"loss": 18.9068,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2457002457002457,
|
| 83 |
+
"grad_norm": 12.625,
|
| 84 |
+
"learning_rate": 4.926108374384237e-05,
|
| 85 |
+
"loss": 17.7765,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2702702702702703,
|
| 90 |
+
"grad_norm": 24.875,
|
| 91 |
+
"learning_rate": 5.41871921182266e-05,
|
| 92 |
+
"loss": 16.5829,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.29484029484029484,
|
| 97 |
+
"grad_norm": 76.5,
|
| 98 |
+
"learning_rate": 5.9113300492610844e-05,
|
| 99 |
+
"loss": 13.7055,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3194103194103194,
|
| 104 |
+
"grad_norm": 52.25,
|
| 105 |
+
"learning_rate": 6.403940886699507e-05,
|
| 106 |
+
"loss": 7.5296,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.343980343980344,
|
| 111 |
+
"grad_norm": 11.9375,
|
| 112 |
+
"learning_rate": 6.896551724137931e-05,
|
| 113 |
+
"loss": 2.7412,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.36855036855036855,
|
| 118 |
+
"grad_norm": 2.984375,
|
| 119 |
+
"learning_rate": 7.389162561576355e-05,
|
| 120 |
+
"loss": 2.1519,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3931203931203931,
|
| 125 |
+
"grad_norm": 2.625,
|
| 126 |
+
"learning_rate": 7.88177339901478e-05,
|
| 127 |
+
"loss": 1.849,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4176904176904177,
|
| 132 |
+
"grad_norm": 3.21875,
|
| 133 |
+
"learning_rate": 8.374384236453202e-05,
|
| 134 |
+
"loss": 1.6847,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.44226044226044225,
|
| 139 |
+
"grad_norm": 2.5625,
|
| 140 |
+
"learning_rate": 8.866995073891627e-05,
|
| 141 |
+
"loss": 1.528,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4668304668304668,
|
| 146 |
+
"grad_norm": 5.65625,
|
| 147 |
+
"learning_rate": 9.35960591133005e-05,
|
| 148 |
+
"loss": 1.4458,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4914004914004914,
|
| 153 |
+
"grad_norm": 1.96875,
|
| 154 |
+
"learning_rate": 9.852216748768474e-05,
|
| 155 |
+
"loss": 1.3912,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.515970515970516,
|
| 160 |
+
"grad_norm": 4.34375,
|
| 161 |
+
"learning_rate": 0.00010344827586206898,
|
| 162 |
+
"loss": 1.3435,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5405405405405406,
|
| 167 |
+
"grad_norm": 1.484375,
|
| 168 |
+
"learning_rate": 0.0001083743842364532,
|
| 169 |
+
"loss": 1.2916,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5651105651105651,
|
| 174 |
+
"grad_norm": 2.828125,
|
| 175 |
+
"learning_rate": 0.00011330049261083743,
|
| 176 |
+
"loss": 1.268,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5896805896805897,
|
| 181 |
+
"grad_norm": 8.75,
|
| 182 |
+
"learning_rate": 0.00011822660098522169,
|
| 183 |
+
"loss": 1.2419,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6142506142506142,
|
| 188 |
+
"grad_norm": 9.875,
|
| 189 |
+
"learning_rate": 0.00012315270935960593,
|
| 190 |
+
"loss": 1.1848,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6388206388206388,
|
| 195 |
+
"grad_norm": 2.859375,
|
| 196 |
+
"learning_rate": 0.00012807881773399014,
|
| 197 |
+
"loss": 1.1821,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6633906633906634,
|
| 202 |
+
"grad_norm": 2.203125,
|
| 203 |
+
"learning_rate": 0.00013300492610837438,
|
| 204 |
+
"loss": 1.1308,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.687960687960688,
|
| 209 |
+
"grad_norm": 3.453125,
|
| 210 |
+
"learning_rate": 0.00013793103448275863,
|
| 211 |
+
"loss": 1.1049,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7125307125307125,
|
| 216 |
+
"grad_norm": 5.53125,
|
| 217 |
+
"learning_rate": 0.00014285714285714287,
|
| 218 |
+
"loss": 1.0904,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7371007371007371,
|
| 223 |
+
"grad_norm": 36.75,
|
| 224 |
+
"learning_rate": 0.0001477832512315271,
|
| 225 |
+
"loss": 1.0884,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7616707616707616,
|
| 230 |
+
"grad_norm": 2.1875,
|
| 231 |
+
"learning_rate": 0.00015270935960591132,
|
| 232 |
+
"loss": 1.0498,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7862407862407862,
|
| 237 |
+
"grad_norm": 3.734375,
|
| 238 |
+
"learning_rate": 0.0001576354679802956,
|
| 239 |
+
"loss": 1.0656,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8108108108108109,
|
| 244 |
+
"grad_norm": 2.71875,
|
| 245 |
+
"learning_rate": 0.0001625615763546798,
|
| 246 |
+
"loss": 1.0436,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8353808353808354,
|
| 251 |
+
"grad_norm": 3.03125,
|
| 252 |
+
"learning_rate": 0.00016748768472906405,
|
| 253 |
+
"loss": 1.0507,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.85995085995086,
|
| 258 |
+
"grad_norm": 2.109375,
|
| 259 |
+
"learning_rate": 0.00017241379310344826,
|
| 260 |
+
"loss": 1.0593,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8845208845208845,
|
| 265 |
+
"grad_norm": 28.875,
|
| 266 |
+
"learning_rate": 0.00017733990147783253,
|
| 267 |
+
"loss": 1.0432,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9090909090909091,
|
| 272 |
+
"grad_norm": 1.4609375,
|
| 273 |
+
"learning_rate": 0.00018226600985221675,
|
| 274 |
+
"loss": 1.0558,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9336609336609336,
|
| 279 |
+
"grad_norm": 3.796875,
|
| 280 |
+
"learning_rate": 0.000187192118226601,
|
| 281 |
+
"loss": 1.0305,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9582309582309583,
|
| 286 |
+
"grad_norm": 1.9140625,
|
| 287 |
+
"learning_rate": 0.00019211822660098523,
|
| 288 |
+
"loss": 1.0315,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9828009828009828,
|
| 293 |
+
"grad_norm": 3.28125,
|
| 294 |
+
"learning_rate": 0.00019704433497536947,
|
| 295 |
+
"loss": 1.0031,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.9975429975429976,
|
| 300 |
+
"eval_loss": 2.529916286468506,
|
| 301 |
+
"eval_runtime": 2.0529,
|
| 302 |
+
"eval_samples_per_second": 4.871,
|
| 303 |
+
"eval_steps_per_second": 0.974,
|
| 304 |
+
"step": 203
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.0073710073710074,
|
| 308 |
+
"grad_norm": 9.4375,
|
| 309 |
+
"learning_rate": 0.00019999940863962815,
|
| 310 |
+
"loss": 1.0212,
|
| 311 |
+
"step": 205
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.031941031941032,
|
| 315 |
+
"grad_norm": 1.421875,
|
| 316 |
+
"learning_rate": 0.00019999275591576766,
|
| 317 |
+
"loss": 0.9729,
|
| 318 |
+
"step": 210
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.0565110565110565,
|
| 322 |
+
"grad_norm": 5.875,
|
| 323 |
+
"learning_rate": 0.00019997871176098827,
|
| 324 |
+
"loss": 0.9684,
|
| 325 |
+
"step": 215
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.0810810810810811,
|
| 329 |
+
"grad_norm": 3.328125,
|
| 330 |
+
"learning_rate": 0.00019995727721342914,
|
| 331 |
+
"loss": 0.9533,
|
| 332 |
+
"step": 220
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.1056511056511056,
|
| 336 |
+
"grad_norm": 1.1953125,
|
| 337 |
+
"learning_rate": 0.00019992845385752485,
|
| 338 |
+
"loss": 0.9418,
|
| 339 |
+
"step": 225
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.1302211302211302,
|
| 343 |
+
"grad_norm": 2.3125,
|
| 344 |
+
"learning_rate": 0.00019989224382388813,
|
| 345 |
+
"loss": 0.9365,
|
| 346 |
+
"step": 230
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.154791154791155,
|
| 350 |
+
"grad_norm": 27.75,
|
| 351 |
+
"learning_rate": 0.00019984864978915253,
|
| 352 |
+
"loss": 0.9535,
|
| 353 |
+
"step": 235
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.1793611793611793,
|
| 357 |
+
"grad_norm": 3.03125,
|
| 358 |
+
"learning_rate": 0.00019979767497577445,
|
| 359 |
+
"loss": 0.9555,
|
| 360 |
+
"step": 240
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.203931203931204,
|
| 364 |
+
"grad_norm": 3.34375,
|
| 365 |
+
"learning_rate": 0.000199739323151795,
|
| 366 |
+
"loss": 0.9599,
|
| 367 |
+
"step": 245
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.2285012285012284,
|
| 371 |
+
"grad_norm": 1.8671875,
|
| 372 |
+
"learning_rate": 0.00019967359863056134,
|
| 373 |
+
"loss": 0.9268,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.253071253071253,
|
| 378 |
+
"grad_norm": 1.96875,
|
| 379 |
+
"learning_rate": 0.00019960050627040806,
|
| 380 |
+
"loss": 0.9183,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.2776412776412776,
|
| 385 |
+
"grad_norm": 8.125,
|
| 386 |
+
"learning_rate": 0.0001995200514742978,
|
| 387 |
+
"loss": 0.9041,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.3022113022113022,
|
| 392 |
+
"grad_norm": 0.8046875,
|
| 393 |
+
"learning_rate": 0.0001994322401894221,
|
| 394 |
+
"loss": 0.915,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.3267813267813269,
|
| 399 |
+
"grad_norm": 0.96484375,
|
| 400 |
+
"learning_rate": 0.00019933707890676158,
|
| 401 |
+
"loss": 0.9009,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.3513513513513513,
|
| 406 |
+
"grad_norm": 3.515625,
|
| 407 |
+
"learning_rate": 0.00019923457466060636,
|
| 408 |
+
"loss": 0.918,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.375921375921376,
|
| 413 |
+
"grad_norm": 1.1640625,
|
| 414 |
+
"learning_rate": 0.00019912473502803582,
|
| 415 |
+
"loss": 0.9329,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.4004914004914004,
|
| 420 |
+
"grad_norm": 0.859375,
|
| 421 |
+
"learning_rate": 0.0001990075681283587,
|
| 422 |
+
"loss": 0.9269,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.425061425061425,
|
| 427 |
+
"grad_norm": 4.875,
|
| 428 |
+
"learning_rate": 0.00019888308262251285,
|
| 429 |
+
"loss": 0.9245,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.4496314496314495,
|
| 434 |
+
"grad_norm": 1.3203125,
|
| 435 |
+
"learning_rate": 0.00019875128771242506,
|
| 436 |
+
"loss": 0.9252,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.4742014742014742,
|
| 441 |
+
"grad_norm": 2.203125,
|
| 442 |
+
"learning_rate": 0.00019861219314033077,
|
| 443 |
+
"loss": 0.9055,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.4987714987714988,
|
| 448 |
+
"grad_norm": 2.640625,
|
| 449 |
+
"learning_rate": 0.000198465809188054,
|
| 450 |
+
"loss": 0.9095,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.5233415233415233,
|
| 455 |
+
"grad_norm": 1.7421875,
|
| 456 |
+
"learning_rate": 0.0001983121466762474,
|
| 457 |
+
"loss": 0.9061,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.547911547911548,
|
| 462 |
+
"grad_norm": 3.671875,
|
| 463 |
+
"learning_rate": 0.00019815121696359212,
|
| 464 |
+
"loss": 0.8923,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.5724815724815726,
|
| 469 |
+
"grad_norm": 2.875,
|
| 470 |
+
"learning_rate": 0.00019798303194595846,
|
| 471 |
+
"loss": 0.9021,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.597051597051597,
|
| 476 |
+
"grad_norm": 1.3671875,
|
| 477 |
+
"learning_rate": 0.00019780760405552645,
|
| 478 |
+
"loss": 0.8921,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.6216216216216215,
|
| 483 |
+
"grad_norm": 1.09375,
|
| 484 |
+
"learning_rate": 0.00019762494625986677,
|
| 485 |
+
"loss": 0.8885,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.6461916461916462,
|
| 490 |
+
"grad_norm": 2.671875,
|
| 491 |
+
"learning_rate": 0.00019743507206098233,
|
| 492 |
+
"loss": 0.8895,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.6707616707616708,
|
| 497 |
+
"grad_norm": 0.7265625,
|
| 498 |
+
"learning_rate": 0.00019723799549431007,
|
| 499 |
+
"loss": 0.8976,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.6953316953316953,
|
| 504 |
+
"grad_norm": 1.109375,
|
| 505 |
+
"learning_rate": 0.00019703373112768365,
|
| 506 |
+
"loss": 0.8836,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.71990171990172,
|
| 511 |
+
"grad_norm": 1.0234375,
|
| 512 |
+
"learning_rate": 0.00019682229406025635,
|
| 513 |
+
"loss": 0.8801,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.7444717444717446,
|
| 518 |
+
"grad_norm": 1.6875,
|
| 519 |
+
"learning_rate": 0.00019660369992138517,
|
| 520 |
+
"loss": 0.9048,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.769041769041769,
|
| 525 |
+
"grad_norm": 1.4921875,
|
| 526 |
+
"learning_rate": 0.0001963779648694754,
|
| 527 |
+
"loss": 0.8906,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.7936117936117935,
|
| 532 |
+
"grad_norm": 1.3203125,
|
| 533 |
+
"learning_rate": 0.00019614510559078625,
|
| 534 |
+
"loss": 0.8805,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.8181818181818183,
|
| 539 |
+
"grad_norm": 0.80859375,
|
| 540 |
+
"learning_rate": 0.00019590513929819734,
|
| 541 |
+
"loss": 0.8839,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.8427518427518428,
|
| 546 |
+
"grad_norm": 1.2109375,
|
| 547 |
+
"learning_rate": 0.0001956580837299364,
|
| 548 |
+
"loss": 0.8715,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.8673218673218672,
|
| 553 |
+
"grad_norm": 1.0078125,
|
| 554 |
+
"learning_rate": 0.000195403957148268,
|
| 555 |
+
"loss": 0.8952,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.8918918918918919,
|
| 560 |
+
"grad_norm": 0.75390625,
|
| 561 |
+
"learning_rate": 0.0001951427783381437,
|
| 562 |
+
"loss": 0.8585,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.9164619164619165,
|
| 567 |
+
"grad_norm": 0.953125,
|
| 568 |
+
"learning_rate": 0.0001948745666058134,
|
| 569 |
+
"loss": 0.8868,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.941031941031941,
|
| 574 |
+
"grad_norm": 2.734375,
|
| 575 |
+
"learning_rate": 0.00019459934177739813,
|
| 576 |
+
"loss": 0.8778,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.9656019656019657,
|
| 581 |
+
"grad_norm": 1.8359375,
|
| 582 |
+
"learning_rate": 0.00019431712419742484,
|
| 583 |
+
"loss": 0.8902,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.9901719901719903,
|
| 588 |
+
"grad_norm": 3.078125,
|
| 589 |
+
"learning_rate": 0.00019402793472732217,
|
| 590 |
+
"loss": 0.8685,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.0,
|
| 595 |
+
"eval_loss": 2.4260478019714355,
|
| 596 |
+
"eval_runtime": 2.0431,
|
| 597 |
+
"eval_samples_per_second": 4.895,
|
| 598 |
+
"eval_steps_per_second": 0.979,
|
| 599 |
+
"step": 407
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.0147420147420148,
|
| 603 |
+
"grad_norm": 1.359375,
|
| 604 |
+
"learning_rate": 0.00019373179474387858,
|
| 605 |
+
"loss": 0.8187,
|
| 606 |
+
"step": 410
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.039312039312039,
|
| 610 |
+
"grad_norm": 0.9140625,
|
| 611 |
+
"learning_rate": 0.0001934287261376622,
|
| 612 |
+
"loss": 0.7808,
|
| 613 |
+
"step": 415
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.063882063882064,
|
| 617 |
+
"grad_norm": 0.73828125,
|
| 618 |
+
"learning_rate": 0.00019311875131140246,
|
| 619 |
+
"loss": 0.7746,
|
| 620 |
+
"step": 420
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.0884520884520885,
|
| 624 |
+
"grad_norm": 0.71875,
|
| 625 |
+
"learning_rate": 0.00019280189317833445,
|
| 626 |
+
"loss": 0.7761,
|
| 627 |
+
"step": 425
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.113022113022113,
|
| 631 |
+
"grad_norm": 1.5859375,
|
| 632 |
+
"learning_rate": 0.00019247817516050483,
|
| 633 |
+
"loss": 0.7781,
|
| 634 |
+
"step": 430
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.1375921375921374,
|
| 638 |
+
"grad_norm": 0.93359375,
|
| 639 |
+
"learning_rate": 0.00019214762118704076,
|
| 640 |
+
"loss": 0.7648,
|
| 641 |
+
"step": 435
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.1621621621621623,
|
| 645 |
+
"grad_norm": 0.828125,
|
| 646 |
+
"learning_rate": 0.0001918102556923809,
|
| 647 |
+
"loss": 0.7926,
|
| 648 |
+
"step": 440
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.1867321867321867,
|
| 652 |
+
"grad_norm": 2.0625,
|
| 653 |
+
"learning_rate": 0.0001914661036144692,
|
| 654 |
+
"loss": 0.7821,
|
| 655 |
+
"step": 445
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.211302211302211,
|
| 659 |
+
"grad_norm": 1.0234375,
|
| 660 |
+
"learning_rate": 0.00019111519039291167,
|
| 661 |
+
"loss": 0.7807,
|
| 662 |
+
"step": 450
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.235872235872236,
|
| 666 |
+
"grad_norm": 1.484375,
|
| 667 |
+
"learning_rate": 0.00019075754196709572,
|
| 668 |
+
"loss": 0.7994,
|
| 669 |
+
"step": 455
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.2604422604422605,
|
| 673 |
+
"grad_norm": 1.0625,
|
| 674 |
+
"learning_rate": 0.0001903931847742728,
|
| 675 |
+
"loss": 0.7777,
|
| 676 |
+
"step": 460
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.285012285012285,
|
| 680 |
+
"grad_norm": 1.8203125,
|
| 681 |
+
"learning_rate": 0.00019002214574760423,
|
| 682 |
+
"loss": 0.7781,
|
| 683 |
+
"step": 465
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.30958230958231,
|
| 687 |
+
"grad_norm": 1.484375,
|
| 688 |
+
"learning_rate": 0.0001896444523141701,
|
| 689 |
+
"loss": 0.7977,
|
| 690 |
+
"step": 470
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.3341523341523343,
|
| 694 |
+
"grad_norm": 0.73046875,
|
| 695 |
+
"learning_rate": 0.00018926013239294216,
|
| 696 |
+
"loss": 0.7758,
|
| 697 |
+
"step": 475
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.3587223587223587,
|
| 701 |
+
"grad_norm": 1.0234375,
|
| 702 |
+
"learning_rate": 0.00018886921439271984,
|
| 703 |
+
"loss": 0.8019,
|
| 704 |
+
"step": 480
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.383292383292383,
|
| 708 |
+
"grad_norm": 0.91015625,
|
| 709 |
+
"learning_rate": 0.00018847172721003043,
|
| 710 |
+
"loss": 0.7829,
|
| 711 |
+
"step": 485
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.407862407862408,
|
| 715 |
+
"grad_norm": 0.984375,
|
| 716 |
+
"learning_rate": 0.00018806770022699278,
|
| 717 |
+
"loss": 0.7759,
|
| 718 |
+
"step": 490
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.4324324324324325,
|
| 722 |
+
"grad_norm": 0.8828125,
|
| 723 |
+
"learning_rate": 0.0001876571633091458,
|
| 724 |
+
"loss": 0.776,
|
| 725 |
+
"step": 495
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.457002457002457,
|
| 729 |
+
"grad_norm": 1.1875,
|
| 730 |
+
"learning_rate": 0.00018724014680324057,
|
| 731 |
+
"loss": 0.7885,
|
| 732 |
+
"step": 500
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.4815724815724813,
|
| 736 |
+
"grad_norm": 0.9375,
|
| 737 |
+
"learning_rate": 0.00018681668153499697,
|
| 738 |
+
"loss": 0.7929,
|
| 739 |
+
"step": 505
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.506142506142506,
|
| 743 |
+
"grad_norm": 4.25,
|
| 744 |
+
"learning_rate": 0.00018638679880682543,
|
| 745 |
+
"loss": 0.7835,
|
| 746 |
+
"step": 510
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.5307125307125307,
|
| 750 |
+
"grad_norm": 1.109375,
|
| 751 |
+
"learning_rate": 0.00018595053039551274,
|
| 752 |
+
"loss": 0.7981,
|
| 753 |
+
"step": 515
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.555282555282555,
|
| 757 |
+
"grad_norm": 0.80078125,
|
| 758 |
+
"learning_rate": 0.00018550790854987323,
|
| 759 |
+
"loss": 0.7831,
|
| 760 |
+
"step": 520
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.57985257985258,
|
| 764 |
+
"grad_norm": 0.92578125,
|
| 765 |
+
"learning_rate": 0.00018505896598836508,
|
| 766 |
+
"loss": 0.7863,
|
| 767 |
+
"step": 525
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 2.6044226044226044,
|
| 771 |
+
"grad_norm": 0.84375,
|
| 772 |
+
"learning_rate": 0.00018460373589667154,
|
| 773 |
+
"loss": 0.7929,
|
| 774 |
+
"step": 530
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.628992628992629,
|
| 778 |
+
"grad_norm": 1.1015625,
|
| 779 |
+
"learning_rate": 0.00018414225192524806,
|
| 780 |
+
"loss": 0.7722,
|
| 781 |
+
"step": 535
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 2.6535626535626538,
|
| 785 |
+
"grad_norm": 1.390625,
|
| 786 |
+
"learning_rate": 0.00018367454818683473,
|
| 787 |
+
"loss": 0.7832,
|
| 788 |
+
"step": 540
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.678132678132678,
|
| 792 |
+
"grad_norm": 1.2890625,
|
| 793 |
+
"learning_rate": 0.00018320065925393468,
|
| 794 |
+
"loss": 0.7746,
|
| 795 |
+
"step": 545
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 2.7027027027027026,
|
| 799 |
+
"grad_norm": 1.59375,
|
| 800 |
+
"learning_rate": 0.00018272062015625872,
|
| 801 |
+
"loss": 0.7826,
|
| 802 |
+
"step": 550
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.7272727272727275,
|
| 806 |
+
"grad_norm": 1.328125,
|
| 807 |
+
"learning_rate": 0.0001822344663781356,
|
| 808 |
+
"loss": 0.7946,
|
| 809 |
+
"step": 555
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.751842751842752,
|
| 813 |
+
"grad_norm": 1.0234375,
|
| 814 |
+
"learning_rate": 0.00018174223385588917,
|
| 815 |
+
"loss": 0.7849,
|
| 816 |
+
"step": 560
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.7764127764127764,
|
| 820 |
+
"grad_norm": 10.3125,
|
| 821 |
+
"learning_rate": 0.00018124395897518224,
|
| 822 |
+
"loss": 0.7783,
|
| 823 |
+
"step": 565
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 2.800982800982801,
|
| 827 |
+
"grad_norm": 0.8515625,
|
| 828 |
+
"learning_rate": 0.0001807396785683264,
|
| 829 |
+
"loss": 0.8005,
|
| 830 |
+
"step": 570
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.8255528255528253,
|
| 834 |
+
"grad_norm": 0.83984375,
|
| 835 |
+
"learning_rate": 0.00018022942991156,
|
| 836 |
+
"loss": 0.7981,
|
| 837 |
+
"step": 575
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.85012285012285,
|
| 841 |
+
"grad_norm": 1.015625,
|
| 842 |
+
"learning_rate": 0.00017971325072229226,
|
| 843 |
+
"loss": 0.7887,
|
| 844 |
+
"step": 580
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.8746928746928746,
|
| 848 |
+
"grad_norm": 0.8125,
|
| 849 |
+
"learning_rate": 0.0001791911791563154,
|
| 850 |
+
"loss": 0.8013,
|
| 851 |
+
"step": 585
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 2.899262899262899,
|
| 855 |
+
"grad_norm": 1.2578125,
|
| 856 |
+
"learning_rate": 0.00017866325380498416,
|
| 857 |
+
"loss": 0.777,
|
| 858 |
+
"step": 590
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.923832923832924,
|
| 862 |
+
"grad_norm": 1.03125,
|
| 863 |
+
"learning_rate": 0.00017812951369236316,
|
| 864 |
+
"loss": 0.7992,
|
| 865 |
+
"step": 595
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.9484029484029484,
|
| 869 |
+
"grad_norm": 1.640625,
|
| 870 |
+
"learning_rate": 0.00017758999827234212,
|
| 871 |
+
"loss": 0.7856,
|
| 872 |
+
"step": 600
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.972972972972973,
|
| 876 |
+
"grad_norm": 3.109375,
|
| 877 |
+
"learning_rate": 0.00017704474742571969,
|
| 878 |
+
"loss": 0.787,
|
| 879 |
+
"step": 605
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.9975429975429977,
|
| 883 |
+
"grad_norm": 3.8125,
|
| 884 |
+
"learning_rate": 0.00017649380145725517,
|
| 885 |
+
"loss": 0.8,
|
| 886 |
+
"step": 610
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.9975429975429977,
|
| 890 |
+
"eval_loss": 2.5051403045654297,
|
| 891 |
+
"eval_runtime": 2.0509,
|
| 892 |
+
"eval_samples_per_second": 4.876,
|
| 893 |
+
"eval_steps_per_second": 0.975,
|
| 894 |
+
"step": 610
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 3.022113022113022,
|
| 898 |
+
"grad_norm": 1.7890625,
|
| 899 |
+
"learning_rate": 0.00017593720109268944,
|
| 900 |
+
"loss": 0.6916,
|
| 901 |
+
"step": 615
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.0466830466830466,
|
| 905 |
+
"grad_norm": 9.4375,
|
| 906 |
+
"learning_rate": 0.00017537498747573443,
|
| 907 |
+
"loss": 0.6614,
|
| 908 |
+
"step": 620
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 3.0712530712530715,
|
| 912 |
+
"grad_norm": 0.96484375,
|
| 913 |
+
"learning_rate": 0.00017480720216503183,
|
| 914 |
+
"loss": 0.6639,
|
| 915 |
+
"step": 625
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 3.095823095823096,
|
| 919 |
+
"grad_norm": 0.74609375,
|
| 920 |
+
"learning_rate": 0.000174233887131081,
|
| 921 |
+
"loss": 0.6579,
|
| 922 |
+
"step": 630
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 3.1203931203931203,
|
| 926 |
+
"grad_norm": 1.015625,
|
| 927 |
+
"learning_rate": 0.0001736550847531366,
|
| 928 |
+
"loss": 0.6546,
|
| 929 |
+
"step": 635
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.1449631449631448,
|
| 933 |
+
"grad_norm": 0.8203125,
|
| 934 |
+
"learning_rate": 0.00017307083781607595,
|
| 935 |
+
"loss": 0.6731,
|
| 936 |
+
"step": 640
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.1695331695331697,
|
| 940 |
+
"grad_norm": 0.87890625,
|
| 941 |
+
"learning_rate": 0.00017248118950723634,
|
| 942 |
+
"loss": 0.6761,
|
| 943 |
+
"step": 645
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.194103194103194,
|
| 947 |
+
"grad_norm": 0.98828125,
|
| 948 |
+
"learning_rate": 0.00017188618341322254,
|
| 949 |
+
"loss": 0.6761,
|
| 950 |
+
"step": 650
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.2186732186732185,
|
| 954 |
+
"grad_norm": 1.7421875,
|
| 955 |
+
"learning_rate": 0.00017128586351668524,
|
| 956 |
+
"loss": 0.666,
|
| 957 |
+
"step": 655
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 3.2432432432432434,
|
| 961 |
+
"grad_norm": 1.3515625,
|
| 962 |
+
"learning_rate": 0.00017068027419306936,
|
| 963 |
+
"loss": 0.6677,
|
| 964 |
+
"step": 660
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 3.267813267813268,
|
| 968 |
+
"grad_norm": 0.9609375,
|
| 969 |
+
"learning_rate": 0.00017006946020733425,
|
| 970 |
+
"loss": 0.6663,
|
| 971 |
+
"step": 665
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.2923832923832923,
|
| 975 |
+
"grad_norm": 0.77734375,
|
| 976 |
+
"learning_rate": 0.00016945346671064452,
|
| 977 |
+
"loss": 0.6762,
|
| 978 |
+
"step": 670
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 3.3169533169533167,
|
| 982 |
+
"grad_norm": 0.7578125,
|
| 983 |
+
"learning_rate": 0.00016883233923703248,
|
| 984 |
+
"loss": 0.6842,
|
| 985 |
+
"step": 675
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.3415233415233416,
|
| 989 |
+
"grad_norm": 1.359375,
|
| 990 |
+
"learning_rate": 0.00016820612370003221,
|
| 991 |
+
"loss": 0.6756,
|
| 992 |
+
"step": 680
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.366093366093366,
|
| 996 |
+
"grad_norm": 1.015625,
|
| 997 |
+
"learning_rate": 0.00016757486638928587,
|
| 998 |
+
"loss": 0.6757,
|
| 999 |
+
"step": 685
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 3.3906633906633905,
|
| 1003 |
+
"grad_norm": 1.5390625,
|
| 1004 |
+
"learning_rate": 0.00016693861396712168,
|
| 1005 |
+
"loss": 0.6971,
|
| 1006 |
+
"step": 690
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 3.4152334152334154,
|
| 1010 |
+
"grad_norm": 1.125,
|
| 1011 |
+
"learning_rate": 0.00016629741346510496,
|
| 1012 |
+
"loss": 0.6837,
|
| 1013 |
+
"step": 695
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.43980343980344,
|
| 1017 |
+
"grad_norm": 2.6875,
|
| 1018 |
+
"learning_rate": 0.00016565131228056133,
|
| 1019 |
+
"loss": 0.6836,
|
| 1020 |
+
"step": 700
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 3.4643734643734643,
|
| 1024 |
+
"grad_norm": 0.77734375,
|
| 1025 |
+
"learning_rate": 0.00016500035817307334,
|
| 1026 |
+
"loss": 0.6719,
|
| 1027 |
+
"step": 705
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 3.488943488943489,
|
| 1031 |
+
"grad_norm": 0.75,
|
| 1032 |
+
"learning_rate": 0.0001643445992609498,
|
| 1033 |
+
"loss": 0.6722,
|
| 1034 |
+
"step": 710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.5135135135135136,
|
| 1038 |
+
"grad_norm": 0.82421875,
|
| 1039 |
+
"learning_rate": 0.00016368408401766916,
|
| 1040 |
+
"loss": 0.6843,
|
| 1041 |
+
"step": 715
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 3.538083538083538,
|
| 1045 |
+
"grad_norm": 0.85546875,
|
| 1046 |
+
"learning_rate": 0.0001630188612682963,
|
| 1047 |
+
"loss": 0.6787,
|
| 1048 |
+
"step": 720
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 3.562653562653563,
|
| 1052 |
+
"grad_norm": 0.71875,
|
| 1053 |
+
"learning_rate": 0.00016234898018587337,
|
| 1054 |
+
"loss": 0.6639,
|
| 1055 |
+
"step": 725
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.5872235872235874,
|
| 1059 |
+
"grad_norm": 0.96484375,
|
| 1060 |
+
"learning_rate": 0.00016167449028778484,
|
| 1061 |
+
"loss": 0.6951,
|
| 1062 |
+
"step": 730
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 3.611793611793612,
|
| 1066 |
+
"grad_norm": 0.80859375,
|
| 1067 |
+
"learning_rate": 0.0001609954414320973,
|
| 1068 |
+
"loss": 0.6881,
|
| 1069 |
+
"step": 735
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 3.6363636363636362,
|
| 1073 |
+
"grad_norm": 0.703125,
|
| 1074 |
+
"learning_rate": 0.0001603118838138741,
|
| 1075 |
+
"loss": 0.6761,
|
| 1076 |
+
"step": 740
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 3.6609336609336607,
|
| 1080 |
+
"grad_norm": 0.9609375,
|
| 1081 |
+
"learning_rate": 0.00015962386796146462,
|
| 1082 |
+
"loss": 0.69,
|
| 1083 |
+
"step": 745
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 3.6855036855036856,
|
| 1087 |
+
"grad_norm": 0.79296875,
|
| 1088 |
+
"learning_rate": 0.00015893144473276953,
|
| 1089 |
+
"loss": 0.691,
|
| 1090 |
+
"step": 750
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.71007371007371,
|
| 1094 |
+
"grad_norm": 1.03125,
|
| 1095 |
+
"learning_rate": 0.00015823466531148124,
|
| 1096 |
+
"loss": 0.6843,
|
| 1097 |
+
"step": 755
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.7346437346437344,
|
| 1101 |
+
"grad_norm": 0.765625,
|
| 1102 |
+
"learning_rate": 0.00015753358120330042,
|
| 1103 |
+
"loss": 0.7094,
|
| 1104 |
+
"step": 760
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 3.7592137592137593,
|
| 1108 |
+
"grad_norm": 1.0078125,
|
| 1109 |
+
"learning_rate": 0.00015682824423212877,
|
| 1110 |
+
"loss": 0.6892,
|
| 1111 |
+
"step": 765
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 3.7837837837837838,
|
| 1115 |
+
"grad_norm": 0.9296875,
|
| 1116 |
+
"learning_rate": 0.00015611870653623825,
|
| 1117 |
+
"loss": 0.6929,
|
| 1118 |
+
"step": 770
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 3.808353808353808,
|
| 1122 |
+
"grad_norm": 1.53125,
|
| 1123 |
+
"learning_rate": 0.00015540502056441688,
|
| 1124 |
+
"loss": 0.7022,
|
| 1125 |
+
"step": 775
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 3.832923832923833,
|
| 1129 |
+
"grad_norm": 0.91015625,
|
| 1130 |
+
"learning_rate": 0.00015468723907209193,
|
| 1131 |
+
"loss": 0.703,
|
| 1132 |
+
"step": 780
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.8574938574938575,
|
| 1136 |
+
"grad_norm": 0.80078125,
|
| 1137 |
+
"learning_rate": 0.00015396541511743012,
|
| 1138 |
+
"loss": 0.7027,
|
| 1139 |
+
"step": 785
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 3.882063882063882,
|
| 1143 |
+
"grad_norm": 0.98046875,
|
| 1144 |
+
"learning_rate": 0.00015323960205741561,
|
| 1145 |
+
"loss": 0.6829,
|
| 1146 |
+
"step": 790
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 3.906633906633907,
|
| 1150 |
+
"grad_norm": 0.921875,
|
| 1151 |
+
"learning_rate": 0.00015250985354390596,
|
| 1152 |
+
"loss": 0.6945,
|
| 1153 |
+
"step": 795
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 3.9312039312039313,
|
| 1157 |
+
"grad_norm": 0.98046875,
|
| 1158 |
+
"learning_rate": 0.0001517762235196661,
|
| 1159 |
+
"loss": 0.6966,
|
| 1160 |
+
"step": 800
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 3.9557739557739557,
|
| 1164 |
+
"grad_norm": 0.7890625,
|
| 1165 |
+
"learning_rate": 0.00015103876621438086,
|
| 1166 |
+
"loss": 0.6947,
|
| 1167 |
+
"step": 805
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 3.98034398034398,
|
| 1171 |
+
"grad_norm": 0.89453125,
|
| 1172 |
+
"learning_rate": 0.00015029753614064645,
|
| 1173 |
+
"loss": 0.6938,
|
| 1174 |
+
"step": 810
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 4.0,
|
| 1178 |
+
"eval_loss": 2.6558120250701904,
|
| 1179 |
+
"eval_runtime": 2.0446,
|
| 1180 |
+
"eval_samples_per_second": 4.891,
|
| 1181 |
+
"eval_steps_per_second": 0.978,
|
| 1182 |
+
"step": 814
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 4.004914004914005,
|
| 1186 |
+
"grad_norm": 1.5703125,
|
| 1187 |
+
"learning_rate": 0.00014955258808994096,
|
| 1188 |
+
"loss": 0.6676,
|
| 1189 |
+
"step": 815
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 4.0294840294840295,
|
| 1193 |
+
"grad_norm": 0.92578125,
|
| 1194 |
+
"learning_rate": 0.00014880397712857386,
|
| 1195 |
+
"loss": 0.5659,
|
| 1196 |
+
"step": 820
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.054054054054054,
|
| 1200 |
+
"grad_norm": 0.7890625,
|
| 1201 |
+
"learning_rate": 0.00014805175859361594,
|
| 1202 |
+
"loss": 0.5525,
|
| 1203 |
+
"step": 825
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.078624078624078,
|
| 1207 |
+
"grad_norm": 1.328125,
|
| 1208 |
+
"learning_rate": 0.00014729598808880861,
|
| 1209 |
+
"loss": 0.5546,
|
| 1210 |
+
"step": 830
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 4.103194103194103,
|
| 1214 |
+
"grad_norm": 0.96875,
|
| 1215 |
+
"learning_rate": 0.00014653672148045357,
|
| 1216 |
+
"loss": 0.5665,
|
| 1217 |
+
"step": 835
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.127764127764128,
|
| 1221 |
+
"grad_norm": 1.0078125,
|
| 1222 |
+
"learning_rate": 0.00014577401489328335,
|
| 1223 |
+
"loss": 0.565,
|
| 1224 |
+
"step": 840
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.152334152334152,
|
| 1228 |
+
"grad_norm": 0.7734375,
|
| 1229 |
+
"learning_rate": 0.0001450079247063127,
|
| 1230 |
+
"loss": 0.5764,
|
| 1231 |
+
"step": 845
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 4.176904176904177,
|
| 1235 |
+
"grad_norm": 0.9375,
|
| 1236 |
+
"learning_rate": 0.00014423850754867075,
|
| 1237 |
+
"loss": 0.5565,
|
| 1238 |
+
"step": 850
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.201474201474202,
|
| 1242 |
+
"grad_norm": 0.921875,
|
| 1243 |
+
"learning_rate": 0.0001434658202954153,
|
| 1244 |
+
"loss": 0.5572,
|
| 1245 |
+
"step": 855
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 4.226044226044226,
|
| 1249 |
+
"grad_norm": 2.0,
|
| 1250 |
+
"learning_rate": 0.00014268992006332846,
|
| 1251 |
+
"loss": 0.5719,
|
| 1252 |
+
"step": 860
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 4.250614250614251,
|
| 1256 |
+
"grad_norm": 0.98828125,
|
| 1257 |
+
"learning_rate": 0.0001419108642066947,
|
| 1258 |
+
"loss": 0.5644,
|
| 1259 |
+
"step": 865
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 4.275184275184275,
|
| 1263 |
+
"grad_norm": 0.88671875,
|
| 1264 |
+
"learning_rate": 0.00014112871031306119,
|
| 1265 |
+
"loss": 0.5785,
|
| 1266 |
+
"step": 870
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 4.2997542997543,
|
| 1270 |
+
"grad_norm": 1.734375,
|
| 1271 |
+
"learning_rate": 0.00014034351619898088,
|
| 1272 |
+
"loss": 0.5825,
|
| 1273 |
+
"step": 875
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 4.324324324324325,
|
| 1277 |
+
"grad_norm": 1.1171875,
|
| 1278 |
+
"learning_rate": 0.00013955533990573886,
|
| 1279 |
+
"loss": 0.5752,
|
| 1280 |
+
"step": 880
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.348894348894349,
|
| 1284 |
+
"grad_norm": 1.1484375,
|
| 1285 |
+
"learning_rate": 0.00013876423969506194,
|
| 1286 |
+
"loss": 0.5863,
|
| 1287 |
+
"step": 885
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 4.3734643734643734,
|
| 1291 |
+
"grad_norm": 1.7421875,
|
| 1292 |
+
"learning_rate": 0.00013797027404481184,
|
| 1293 |
+
"loss": 0.5826,
|
| 1294 |
+
"step": 890
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 4.398034398034398,
|
| 1298 |
+
"grad_norm": 0.99609375,
|
| 1299 |
+
"learning_rate": 0.0001371735016446627,
|
| 1300 |
+
"loss": 0.576,
|
| 1301 |
+
"step": 895
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 4.422604422604422,
|
| 1305 |
+
"grad_norm": 0.875,
|
| 1306 |
+
"learning_rate": 0.00013637398139176255,
|
| 1307 |
+
"loss": 0.577,
|
| 1308 |
+
"step": 900
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 4.447174447174447,
|
| 1312 |
+
"grad_norm": 0.9375,
|
| 1313 |
+
"learning_rate": 0.00013557177238637986,
|
| 1314 |
+
"loss": 0.5832,
|
| 1315 |
+
"step": 905
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 4.471744471744472,
|
| 1319 |
+
"grad_norm": 0.9375,
|
| 1320 |
+
"learning_rate": 0.00013476693392753476,
|
| 1321 |
+
"loss": 0.5856,
|
| 1322 |
+
"step": 910
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.496314496314496,
|
| 1326 |
+
"grad_norm": 1.1328125,
|
| 1327 |
+
"learning_rate": 0.00013395952550861572,
|
| 1328 |
+
"loss": 0.592,
|
| 1329 |
+
"step": 915
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 4.520884520884521,
|
| 1333 |
+
"grad_norm": 0.97265625,
|
| 1334 |
+
"learning_rate": 0.00013314960681298175,
|
| 1335 |
+
"loss": 0.5861,
|
| 1336 |
+
"step": 920
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 4.545454545454545,
|
| 1340 |
+
"grad_norm": 0.97265625,
|
| 1341 |
+
"learning_rate": 0.0001323372377095507,
|
| 1342 |
+
"loss": 0.5814,
|
| 1343 |
+
"step": 925
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 4.57002457002457,
|
| 1347 |
+
"grad_norm": 0.91796875,
|
| 1348 |
+
"learning_rate": 0.0001315224782483737,
|
| 1349 |
+
"loss": 0.5847,
|
| 1350 |
+
"step": 930
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 4.594594594594595,
|
| 1354 |
+
"grad_norm": 0.8203125,
|
| 1355 |
+
"learning_rate": 0.00013070538865619642,
|
| 1356 |
+
"loss": 0.5773,
|
| 1357 |
+
"step": 935
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 4.61916461916462,
|
| 1361 |
+
"grad_norm": 0.859375,
|
| 1362 |
+
"learning_rate": 0.00012988602933200689,
|
| 1363 |
+
"loss": 0.5723,
|
| 1364 |
+
"step": 940
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.643734643734644,
|
| 1368 |
+
"grad_norm": 1.1171875,
|
| 1369 |
+
"learning_rate": 0.0001290644608425711,
|
| 1370 |
+
"loss": 0.5918,
|
| 1371 |
+
"step": 945
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 4.6683046683046685,
|
| 1375 |
+
"grad_norm": 0.90234375,
|
| 1376 |
+
"learning_rate": 0.0001282407439179557,
|
| 1377 |
+
"loss": 0.5932,
|
| 1378 |
+
"step": 950
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 4.6928746928746925,
|
| 1382 |
+
"grad_norm": 0.8671875,
|
| 1383 |
+
"learning_rate": 0.00012741493944703905,
|
| 1384 |
+
"loss": 0.5868,
|
| 1385 |
+
"step": 955
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.717444717444717,
|
| 1389 |
+
"grad_norm": 0.83203125,
|
| 1390 |
+
"learning_rate": 0.0001265871084730101,
|
| 1391 |
+
"loss": 0.5944,
|
| 1392 |
+
"step": 960
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 4.742014742014742,
|
| 1396 |
+
"grad_norm": 0.9765625,
|
| 1397 |
+
"learning_rate": 0.00012575731218885625,
|
| 1398 |
+
"loss": 0.5893,
|
| 1399 |
+
"step": 965
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 4.766584766584766,
|
| 1403 |
+
"grad_norm": 1.140625,
|
| 1404 |
+
"learning_rate": 0.00012492561193284008,
|
| 1405 |
+
"loss": 0.5791,
|
| 1406 |
+
"step": 970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.791154791154791,
|
| 1410 |
+
"grad_norm": 1.1484375,
|
| 1411 |
+
"learning_rate": 0.00012409206918396503,
|
| 1412 |
+
"loss": 0.5876,
|
| 1413 |
+
"step": 975
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 4.815724815724816,
|
| 1417 |
+
"grad_norm": 0.7890625,
|
| 1418 |
+
"learning_rate": 0.00012325674555743106,
|
| 1419 |
+
"loss": 0.5871,
|
| 1420 |
+
"step": 980
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 4.84029484029484,
|
| 1424 |
+
"grad_norm": 0.8046875,
|
| 1425 |
+
"learning_rate": 0.0001224197028000799,
|
| 1426 |
+
"loss": 0.5987,
|
| 1427 |
+
"step": 985
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 4.864864864864865,
|
| 1431 |
+
"grad_norm": 0.92578125,
|
| 1432 |
+
"learning_rate": 0.000121581002785831,
|
| 1433 |
+
"loss": 0.5857,
|
| 1434 |
+
"step": 990
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 4.88943488943489,
|
| 1438 |
+
"grad_norm": 0.80078125,
|
| 1439 |
+
"learning_rate": 0.00012074070751110751,
|
| 1440 |
+
"loss": 0.5932,
|
| 1441 |
+
"step": 995
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 4.914004914004914,
|
| 1445 |
+
"grad_norm": 1.0625,
|
| 1446 |
+
"learning_rate": 0.00011989887909025388,
|
| 1447 |
+
"loss": 0.5891,
|
| 1448 |
+
"step": 1000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.938574938574939,
|
| 1452 |
+
"grad_norm": 0.796875,
|
| 1453 |
+
"learning_rate": 0.00011905557975094406,
|
| 1454 |
+
"loss": 0.5881,
|
| 1455 |
+
"step": 1005
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 4.963144963144963,
|
| 1459 |
+
"grad_norm": 0.87890625,
|
| 1460 |
+
"learning_rate": 0.00011821087182958186,
|
| 1461 |
+
"loss": 0.5845,
|
| 1462 |
+
"step": 1010
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.987714987714988,
|
| 1466 |
+
"grad_norm": 0.828125,
|
| 1467 |
+
"learning_rate": 0.00011736481776669306,
|
| 1468 |
+
"loss": 0.5865,
|
| 1469 |
+
"step": 1015
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 4.997542997542998,
|
| 1473 |
+
"eval_loss": 2.9162657260894775,
|
| 1474 |
+
"eval_runtime": 2.0535,
|
| 1475 |
+
"eval_samples_per_second": 4.87,
|
| 1476 |
+
"eval_steps_per_second": 0.974,
|
| 1477 |
+
"step": 1017
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 5.012285012285012,
|
| 1481 |
+
"grad_norm": 0.75390625,
|
| 1482 |
+
"learning_rate": 0.0001165174801023096,
|
| 1483 |
+
"loss": 0.5388,
|
| 1484 |
+
"step": 1020
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 5.036855036855036,
|
| 1488 |
+
"grad_norm": 0.953125,
|
| 1489 |
+
"learning_rate": 0.00011566892147134705,
|
| 1490 |
+
"loss": 0.4596,
|
| 1491 |
+
"step": 1025
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 5.061425061425061,
|
| 1495 |
+
"grad_norm": 0.7890625,
|
| 1496 |
+
"learning_rate": 0.00011481920459897417,
|
| 1497 |
+
"loss": 0.4681,
|
| 1498 |
+
"step": 1030
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 5.085995085995086,
|
| 1502 |
+
"grad_norm": 0.8828125,
|
| 1503 |
+
"learning_rate": 0.00011396839229597674,
|
| 1504 |
+
"loss": 0.4716,
|
| 1505 |
+
"step": 1035
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 5.11056511056511,
|
| 1509 |
+
"grad_norm": 0.80078125,
|
| 1510 |
+
"learning_rate": 0.00011311654745411425,
|
| 1511 |
+
"loss": 0.4667,
|
| 1512 |
+
"step": 1040
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 5.135135135135135,
|
| 1516 |
+
"grad_norm": 0.90625,
|
| 1517 |
+
"learning_rate": 0.00011226373304147123,
|
| 1518 |
+
"loss": 0.465,
|
| 1519 |
+
"step": 1045
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 5.15970515970516,
|
| 1523 |
+
"grad_norm": 0.828125,
|
| 1524 |
+
"learning_rate": 0.00011141001209780249,
|
| 1525 |
+
"loss": 0.4751,
|
| 1526 |
+
"step": 1050
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 5.184275184275184,
|
| 1530 |
+
"grad_norm": 1.0390625,
|
| 1531 |
+
"learning_rate": 0.00011055544772987335,
|
| 1532 |
+
"loss": 0.4622,
|
| 1533 |
+
"step": 1055
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 5.208845208845209,
|
| 1537 |
+
"grad_norm": 1.234375,
|
| 1538 |
+
"learning_rate": 0.0001097001031067947,
|
| 1539 |
+
"loss": 0.4839,
|
| 1540 |
+
"step": 1060
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 5.233415233415234,
|
| 1544 |
+
"grad_norm": 0.91015625,
|
| 1545 |
+
"learning_rate": 0.00010884404145535372,
|
| 1546 |
+
"loss": 0.4679,
|
| 1547 |
+
"step": 1065
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.257985257985258,
|
| 1551 |
+
"grad_norm": 0.88671875,
|
| 1552 |
+
"learning_rate": 0.00010798732605534006,
|
| 1553 |
+
"loss": 0.4764,
|
| 1554 |
+
"step": 1070
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 5.282555282555283,
|
| 1558 |
+
"grad_norm": 0.90234375,
|
| 1559 |
+
"learning_rate": 0.00010713002023486816,
|
| 1560 |
+
"loss": 0.4834,
|
| 1561 |
+
"step": 1075
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 5.3071253071253075,
|
| 1565 |
+
"grad_norm": 0.98046875,
|
| 1566 |
+
"learning_rate": 0.00010627218736569624,
|
| 1567 |
+
"loss": 0.4853,
|
| 1568 |
+
"step": 1080
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 5.3316953316953315,
|
| 1572 |
+
"grad_norm": 0.875,
|
| 1573 |
+
"learning_rate": 0.00010541389085854176,
|
| 1574 |
+
"loss": 0.4823,
|
| 1575 |
+
"step": 1085
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 5.356265356265356,
|
| 1579 |
+
"grad_norm": 1.03125,
|
| 1580 |
+
"learning_rate": 0.00010455519415839415,
|
| 1581 |
+
"loss": 0.4838,
|
| 1582 |
+
"step": 1090
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 5.38083538083538,
|
| 1586 |
+
"grad_norm": 0.92578125,
|
| 1587 |
+
"learning_rate": 0.00010369616073982491,
|
| 1588 |
+
"loss": 0.4881,
|
| 1589 |
+
"step": 1095
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 5.405405405405405,
|
| 1593 |
+
"grad_norm": 0.88671875,
|
| 1594 |
+
"learning_rate": 0.00010283685410229571,
|
| 1595 |
+
"loss": 0.485,
|
| 1596 |
+
"step": 1100
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 5.42997542997543,
|
| 1600 |
+
"grad_norm": 1.0390625,
|
| 1601 |
+
"learning_rate": 0.00010197733776546447,
|
| 1602 |
+
"loss": 0.4854,
|
| 1603 |
+
"step": 1105
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 5.454545454545454,
|
| 1607 |
+
"grad_norm": 0.9609375,
|
| 1608 |
+
"learning_rate": 0.00010111767526449004,
|
| 1609 |
+
"loss": 0.4851,
|
| 1610 |
+
"step": 1110
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 5.479115479115479,
|
| 1614 |
+
"grad_norm": 1.265625,
|
| 1615 |
+
"learning_rate": 0.00010025793014533558,
|
| 1616 |
+
"loss": 0.4889,
|
| 1617 |
+
"step": 1115
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 5.503685503685504,
|
| 1621 |
+
"grad_norm": 0.87890625,
|
| 1622 |
+
"learning_rate": 9.939816596007146e-05,
|
| 1623 |
+
"loss": 0.4772,
|
| 1624 |
+
"step": 1120
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 5.528255528255528,
|
| 1628 |
+
"grad_norm": 1.1640625,
|
| 1629 |
+
"learning_rate": 9.853844626217737e-05,
|
| 1630 |
+
"loss": 0.4857,
|
| 1631 |
+
"step": 1125
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 5.552825552825553,
|
| 1635 |
+
"grad_norm": 1.0703125,
|
| 1636 |
+
"learning_rate": 9.767883460184443e-05,
|
| 1637 |
+
"loss": 0.4869,
|
| 1638 |
+
"step": 1130
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 5.577395577395578,
|
| 1642 |
+
"grad_norm": 0.88671875,
|
| 1643 |
+
"learning_rate": 9.681939452127784e-05,
|
| 1644 |
+
"loss": 0.4806,
|
| 1645 |
+
"step": 1135
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.601965601965602,
|
| 1649 |
+
"grad_norm": 1.2421875,
|
| 1650 |
+
"learning_rate": 9.596018954999953e-05,
|
| 1651 |
+
"loss": 0.4897,
|
| 1652 |
+
"step": 1140
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.6265356265356266,
|
| 1656 |
+
"grad_norm": 1.015625,
|
| 1657 |
+
"learning_rate": 9.510128320015224e-05,
|
| 1658 |
+
"loss": 0.4749,
|
| 1659 |
+
"step": 1145
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 5.651105651105651,
|
| 1663 |
+
"grad_norm": 0.93359375,
|
| 1664 |
+
"learning_rate": 9.424273896180482e-05,
|
| 1665 |
+
"loss": 0.4837,
|
| 1666 |
+
"step": 1150
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 5.675675675675675,
|
| 1670 |
+
"grad_norm": 0.8984375,
|
| 1671 |
+
"learning_rate": 9.338462029825886e-05,
|
| 1672 |
+
"loss": 0.4928,
|
| 1673 |
+
"step": 1155
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 5.7002457002457,
|
| 1677 |
+
"grad_norm": 0.97265625,
|
| 1678 |
+
"learning_rate": 9.252699064135758e-05,
|
| 1679 |
+
"loss": 0.477,
|
| 1680 |
+
"step": 1160
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 5.724815724815725,
|
| 1684 |
+
"grad_norm": 0.87890625,
|
| 1685 |
+
"learning_rate": 9.166991338679715e-05,
|
| 1686 |
+
"loss": 0.4882,
|
| 1687 |
+
"step": 1165
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 5.749385749385749,
|
| 1691 |
+
"grad_norm": 1.078125,
|
| 1692 |
+
"learning_rate": 9.081345188944019e-05,
|
| 1693 |
+
"loss": 0.4902,
|
| 1694 |
+
"step": 1170
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 5.773955773955774,
|
| 1698 |
+
"grad_norm": 0.97265625,
|
| 1699 |
+
"learning_rate": 8.995766945863277e-05,
|
| 1700 |
+
"loss": 0.4873,
|
| 1701 |
+
"step": 1175
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 5.798525798525798,
|
| 1705 |
+
"grad_norm": 1.0625,
|
| 1706 |
+
"learning_rate": 8.91026293535247e-05,
|
| 1707 |
+
"loss": 0.4864,
|
| 1708 |
+
"step": 1180
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 5.823095823095823,
|
| 1712 |
+
"grad_norm": 1.1953125,
|
| 1713 |
+
"learning_rate": 8.82483947783932e-05,
|
| 1714 |
+
"loss": 0.4782,
|
| 1715 |
+
"step": 1185
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 5.847665847665848,
|
| 1719 |
+
"grad_norm": 0.921875,
|
| 1720 |
+
"learning_rate": 8.739502887797107e-05,
|
| 1721 |
+
"loss": 0.4918,
|
| 1722 |
+
"step": 1190
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 5.872235872235873,
|
| 1726 |
+
"grad_norm": 0.9296875,
|
| 1727 |
+
"learning_rate": 8.654259473277892e-05,
|
| 1728 |
+
"loss": 0.4955,
|
| 1729 |
+
"step": 1195
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 5.896805896805897,
|
| 1733 |
+
"grad_norm": 1.015625,
|
| 1734 |
+
"learning_rate": 8.569115535446228e-05,
|
| 1735 |
+
"loss": 0.4802,
|
| 1736 |
+
"step": 1200
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 5.921375921375922,
|
| 1740 |
+
"grad_norm": 0.8984375,
|
| 1741 |
+
"learning_rate": 8.484077368113399e-05,
|
| 1742 |
+
"loss": 0.4832,
|
| 1743 |
+
"step": 1205
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 5.945945945945946,
|
| 1747 |
+
"grad_norm": 0.87890625,
|
| 1748 |
+
"learning_rate": 8.399151257272156e-05,
|
| 1749 |
+
"loss": 0.4847,
|
| 1750 |
+
"step": 1210
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 5.9705159705159705,
|
| 1754 |
+
"grad_norm": 0.87890625,
|
| 1755 |
+
"learning_rate": 8.314343480632078e-05,
|
| 1756 |
+
"loss": 0.48,
|
| 1757 |
+
"step": 1215
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 5.995085995085995,
|
| 1761 |
+
"grad_norm": 0.87890625,
|
| 1762 |
+
"learning_rate": 8.229660307155518e-05,
|
| 1763 |
+
"loss": 0.4821,
|
| 1764 |
+
"step": 1220
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 6.0,
|
| 1768 |
+
"eval_loss": 3.3285133838653564,
|
| 1769 |
+
"eval_runtime": 2.0506,
|
| 1770 |
+
"eval_samples_per_second": 4.877,
|
| 1771 |
+
"eval_steps_per_second": 0.975,
|
| 1772 |
+
"step": 1221
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 6.019656019656019,
|
| 1776 |
+
"grad_norm": 0.80078125,
|
| 1777 |
+
"learning_rate": 8.145107996594206e-05,
|
| 1778 |
+
"loss": 0.4087,
|
| 1779 |
+
"step": 1225
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 6.044226044226044,
|
| 1783 |
+
"grad_norm": 0.98046875,
|
| 1784 |
+
"learning_rate": 8.060692799026522e-05,
|
| 1785 |
+
"loss": 0.3843,
|
| 1786 |
+
"step": 1230
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 6.068796068796069,
|
| 1790 |
+
"grad_norm": 0.828125,
|
| 1791 |
+
"learning_rate": 7.976420954395518e-05,
|
| 1792 |
+
"loss": 0.3844,
|
| 1793 |
+
"step": 1235
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 6.093366093366093,
|
| 1797 |
+
"grad_norm": 0.984375,
|
| 1798 |
+
"learning_rate": 7.892298692047621e-05,
|
| 1799 |
+
"loss": 0.3909,
|
| 1800 |
+
"step": 1240
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 6.117936117936118,
|
| 1804 |
+
"grad_norm": 1.1171875,
|
| 1805 |
+
"learning_rate": 7.808332230272209e-05,
|
| 1806 |
+
"loss": 0.393,
|
| 1807 |
+
"step": 1245
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 6.142506142506143,
|
| 1811 |
+
"grad_norm": 0.9453125,
|
| 1812 |
+
"learning_rate": 7.724527775841914e-05,
|
| 1813 |
+
"loss": 0.3818,
|
| 1814 |
+
"step": 1250
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.167076167076167,
|
| 1818 |
+
"grad_norm": 0.921875,
|
| 1819 |
+
"learning_rate": 7.64089152355385e-05,
|
| 1820 |
+
"loss": 0.3938,
|
| 1821 |
+
"step": 1255
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 6.191646191646192,
|
| 1825 |
+
"grad_norm": 0.90234375,
|
| 1826 |
+
"learning_rate": 7.55742965577169e-05,
|
| 1827 |
+
"loss": 0.3885,
|
| 1828 |
+
"step": 1260
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 6.216216216216216,
|
| 1832 |
+
"grad_norm": 0.94921875,
|
| 1833 |
+
"learning_rate": 7.474148341968652e-05,
|
| 1834 |
+
"loss": 0.3889,
|
| 1835 |
+
"step": 1265
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 6.240786240786241,
|
| 1839 |
+
"grad_norm": 1.171875,
|
| 1840 |
+
"learning_rate": 7.391053738271466e-05,
|
| 1841 |
+
"loss": 0.3932,
|
| 1842 |
+
"step": 1270
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 6.2653562653562656,
|
| 1846 |
+
"grad_norm": 0.91015625,
|
| 1847 |
+
"learning_rate": 7.308151987005326e-05,
|
| 1848 |
+
"loss": 0.3823,
|
| 1849 |
+
"step": 1275
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 6.2899262899262895,
|
| 1853 |
+
"grad_norm": 0.91015625,
|
| 1854 |
+
"learning_rate": 7.225449216239821e-05,
|
| 1855 |
+
"loss": 0.3857,
|
| 1856 |
+
"step": 1280
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.314496314496314,
|
| 1860 |
+
"grad_norm": 0.8671875,
|
| 1861 |
+
"learning_rate": 7.142951539335981e-05,
|
| 1862 |
+
"loss": 0.3973,
|
| 1863 |
+
"step": 1285
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 6.339066339066339,
|
| 1867 |
+
"grad_norm": 0.91796875,
|
| 1868 |
+
"learning_rate": 7.060665054494362e-05,
|
| 1869 |
+
"loss": 0.3901,
|
| 1870 |
+
"step": 1290
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 6.363636363636363,
|
| 1874 |
+
"grad_norm": 0.89453125,
|
| 1875 |
+
"learning_rate": 6.978595844304271e-05,
|
| 1876 |
+
"loss": 0.386,
|
| 1877 |
+
"step": 1295
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 6.388206388206388,
|
| 1881 |
+
"grad_norm": 1.03125,
|
| 1882 |
+
"learning_rate": 6.89674997529416e-05,
|
| 1883 |
+
"loss": 0.3881,
|
| 1884 |
+
"step": 1300
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 6.412776412776413,
|
| 1888 |
+
"grad_norm": 0.89453125,
|
| 1889 |
+
"learning_rate": 6.815133497483157e-05,
|
| 1890 |
+
"loss": 0.4006,
|
| 1891 |
+
"step": 1305
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 6.437346437346437,
|
| 1895 |
+
"grad_norm": 1.0234375,
|
| 1896 |
+
"learning_rate": 6.733752443933878e-05,
|
| 1897 |
+
"loss": 0.397,
|
| 1898 |
+
"step": 1310
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.461916461916462,
|
| 1902 |
+
"grad_norm": 1.0703125,
|
| 1903 |
+
"learning_rate": 6.65261283030646e-05,
|
| 1904 |
+
"loss": 0.3972,
|
| 1905 |
+
"step": 1315
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 6.486486486486487,
|
| 1909 |
+
"grad_norm": 1.0390625,
|
| 1910 |
+
"learning_rate": 6.571720654413877e-05,
|
| 1911 |
+
"loss": 0.3965,
|
| 1912 |
+
"step": 1320
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 6.511056511056511,
|
| 1916 |
+
"grad_norm": 0.9296875,
|
| 1917 |
+
"learning_rate": 6.491081895778588e-05,
|
| 1918 |
+
"loss": 0.3961,
|
| 1919 |
+
"step": 1325
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 6.535626535626536,
|
| 1923 |
+
"grad_norm": 1.0625,
|
| 1924 |
+
"learning_rate": 6.410702515190543e-05,
|
| 1925 |
+
"loss": 0.3919,
|
| 1926 |
+
"step": 1330
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 6.560196560196561,
|
| 1930 |
+
"grad_norm": 0.9453125,
|
| 1931 |
+
"learning_rate": 6.330588454266542e-05,
|
| 1932 |
+
"loss": 0.3916,
|
| 1933 |
+
"step": 1335
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 6.584766584766585,
|
| 1937 |
+
"grad_norm": 0.9453125,
|
| 1938 |
+
"learning_rate": 6.250745635011048e-05,
|
| 1939 |
+
"loss": 0.4089,
|
| 1940 |
+
"step": 1340
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.6093366093366095,
|
| 1944 |
+
"grad_norm": 0.9140625,
|
| 1945 |
+
"learning_rate": 6.171179959378437e-05,
|
| 1946 |
+
"loss": 0.401,
|
| 1947 |
+
"step": 1345
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 6.6339066339066335,
|
| 1951 |
+
"grad_norm": 0.89453125,
|
| 1952 |
+
"learning_rate": 6.0918973088367116e-05,
|
| 1953 |
+
"loss": 0.3927,
|
| 1954 |
+
"step": 1350
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 6.658476658476658,
|
| 1958 |
+
"grad_norm": 1.0,
|
| 1959 |
+
"learning_rate": 6.012903543932766e-05,
|
| 1960 |
+
"loss": 0.3899,
|
| 1961 |
+
"step": 1355
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 6.683046683046683,
|
| 1965 |
+
"grad_norm": 0.84765625,
|
| 1966 |
+
"learning_rate": 5.934204503859158e-05,
|
| 1967 |
+
"loss": 0.3952,
|
| 1968 |
+
"step": 1360
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 6.707616707616707,
|
| 1972 |
+
"grad_norm": 0.98828125,
|
| 1973 |
+
"learning_rate": 5.8558060060224817e-05,
|
| 1974 |
+
"loss": 0.3917,
|
| 1975 |
+
"step": 1365
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 6.732186732186732,
|
| 1979 |
+
"grad_norm": 0.92578125,
|
| 1980 |
+
"learning_rate": 5.777713845613364e-05,
|
| 1981 |
+
"loss": 0.3878,
|
| 1982 |
+
"step": 1370
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 6.756756756756757,
|
| 1986 |
+
"grad_norm": 0.95703125,
|
| 1987 |
+
"learning_rate": 5.699933795178052e-05,
|
| 1988 |
+
"loss": 0.4066,
|
| 1989 |
+
"step": 1375
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 6.781326781326781,
|
| 1993 |
+
"grad_norm": 0.90625,
|
| 1994 |
+
"learning_rate": 5.622471604191746e-05,
|
| 1995 |
+
"loss": 0.3993,
|
| 1996 |
+
"step": 1380
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 6.805896805896806,
|
| 2000 |
+
"grad_norm": 0.98828125,
|
| 2001 |
+
"learning_rate": 5.545332998633572e-05,
|
| 2002 |
+
"loss": 0.3975,
|
| 2003 |
+
"step": 1385
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 6.830466830466831,
|
| 2007 |
+
"grad_norm": 0.96875,
|
| 2008 |
+
"learning_rate": 5.46852368056334e-05,
|
| 2009 |
+
"loss": 0.4003,
|
| 2010 |
+
"step": 1390
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 6.855036855036855,
|
| 2014 |
+
"grad_norm": 0.828125,
|
| 2015 |
+
"learning_rate": 5.392049327700026e-05,
|
| 2016 |
+
"loss": 0.3978,
|
| 2017 |
+
"step": 1395
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 6.87960687960688,
|
| 2021 |
+
"grad_norm": 0.90625,
|
| 2022 |
+
"learning_rate": 5.3159155930021e-05,
|
| 2023 |
+
"loss": 0.4017,
|
| 2024 |
+
"step": 1400
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 6.9041769041769046,
|
| 2028 |
+
"grad_norm": 0.90625,
|
| 2029 |
+
"learning_rate": 5.2401281042496494e-05,
|
| 2030 |
+
"loss": 0.3992,
|
| 2031 |
+
"step": 1405
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 6.9287469287469285,
|
| 2035 |
+
"grad_norm": 0.95703125,
|
| 2036 |
+
"learning_rate": 5.164692463628378e-05,
|
| 2037 |
+
"loss": 0.3965,
|
| 2038 |
+
"step": 1410
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 6.953316953316953,
|
| 2042 |
+
"grad_norm": 0.9140625,
|
| 2043 |
+
"learning_rate": 5.0896142473154987e-05,
|
| 2044 |
+
"loss": 0.3883,
|
| 2045 |
+
"step": 1415
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 6.977886977886978,
|
| 2049 |
+
"grad_norm": 0.95703125,
|
| 2050 |
+
"learning_rate": 5.014899005067524e-05,
|
| 2051 |
+
"loss": 0.3899,
|
| 2052 |
+
"step": 1420
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 6.997542997542998,
|
| 2056 |
+
"eval_loss": 3.898437023162842,
|
| 2057 |
+
"eval_runtime": 2.0525,
|
| 2058 |
+
"eval_samples_per_second": 4.872,
|
| 2059 |
+
"eval_steps_per_second": 0.974,
|
| 2060 |
+
"step": 1424
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.002457002457002,
|
| 2064 |
+
"grad_norm": 1.0546875,
|
| 2065 |
+
"learning_rate": 4.940552259810063e-05,
|
| 2066 |
+
"loss": 0.3846,
|
| 2067 |
+
"step": 1425
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.027027027027027,
|
| 2071 |
+
"grad_norm": 0.82421875,
|
| 2072 |
+
"learning_rate": 4.866579507229545e-05,
|
| 2073 |
+
"loss": 0.3325,
|
| 2074 |
+
"step": 1430
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.051597051597051,
|
| 2078 |
+
"grad_norm": 1.0,
|
| 2079 |
+
"learning_rate": 4.792986215366976e-05,
|
| 2080 |
+
"loss": 0.3266,
|
| 2081 |
+
"step": 1435
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.076167076167076,
|
| 2085 |
+
"grad_norm": 1.0078125,
|
| 2086 |
+
"learning_rate": 4.7197778242137755e-05,
|
| 2087 |
+
"loss": 0.3295,
|
| 2088 |
+
"step": 1440
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.100737100737101,
|
| 2092 |
+
"grad_norm": 0.80078125,
|
| 2093 |
+
"learning_rate": 4.646959745309609e-05,
|
| 2094 |
+
"loss": 0.3279,
|
| 2095 |
+
"step": 1445
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.125307125307125,
|
| 2099 |
+
"grad_norm": 0.96875,
|
| 2100 |
+
"learning_rate": 4.574537361342407e-05,
|
| 2101 |
+
"loss": 0.324,
|
| 2102 |
+
"step": 1450
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.14987714987715,
|
| 2106 |
+
"grad_norm": 0.87109375,
|
| 2107 |
+
"learning_rate": 4.502516025750455e-05,
|
| 2108 |
+
"loss": 0.319,
|
| 2109 |
+
"step": 1455
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.174447174447175,
|
| 2113 |
+
"grad_norm": 0.8125,
|
| 2114 |
+
"learning_rate": 4.430901062326681e-05,
|
| 2115 |
+
"loss": 0.3242,
|
| 2116 |
+
"step": 1460
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.199017199017199,
|
| 2120 |
+
"grad_norm": 0.8515625,
|
| 2121 |
+
"learning_rate": 4.359697764825123e-05,
|
| 2122 |
+
"loss": 0.3172,
|
| 2123 |
+
"step": 1465
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.223587223587224,
|
| 2127 |
+
"grad_norm": 0.9140625,
|
| 2128 |
+
"learning_rate": 4.288911396569599e-05,
|
| 2129 |
+
"loss": 0.3275,
|
| 2130 |
+
"step": 1470
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.2481572481572485,
|
| 2134 |
+
"grad_norm": 0.8828125,
|
| 2135 |
+
"learning_rate": 4.21854719006467e-05,
|
| 2136 |
+
"loss": 0.332,
|
| 2137 |
+
"step": 1475
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.2727272727272725,
|
| 2141 |
+
"grad_norm": 1.015625,
|
| 2142 |
+
"learning_rate": 4.148610346608837e-05,
|
| 2143 |
+
"loss": 0.3359,
|
| 2144 |
+
"step": 1480
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.297297297297297,
|
| 2148 |
+
"grad_norm": 0.87109375,
|
| 2149 |
+
"learning_rate": 4.079106035910073e-05,
|
| 2150 |
+
"loss": 0.3242,
|
| 2151 |
+
"step": 1485
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.321867321867322,
|
| 2155 |
+
"grad_norm": 0.86328125,
|
| 2156 |
+
"learning_rate": 4.010039395703664e-05,
|
| 2157 |
+
"loss": 0.3273,
|
| 2158 |
+
"step": 1490
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.346437346437346,
|
| 2162 |
+
"grad_norm": 0.9140625,
|
| 2163 |
+
"learning_rate": 3.94141553137245e-05,
|
| 2164 |
+
"loss": 0.3274,
|
| 2165 |
+
"step": 1495
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.371007371007371,
|
| 2169 |
+
"grad_norm": 0.88671875,
|
| 2170 |
+
"learning_rate": 3.873239515569429e-05,
|
| 2171 |
+
"loss": 0.3266,
|
| 2172 |
+
"step": 1500
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.395577395577396,
|
| 2176 |
+
"grad_norm": 0.84765625,
|
| 2177 |
+
"learning_rate": 3.80551638784277e-05,
|
| 2178 |
+
"loss": 0.3363,
|
| 2179 |
+
"step": 1505
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.42014742014742,
|
| 2183 |
+
"grad_norm": 0.87109375,
|
| 2184 |
+
"learning_rate": 3.738251154263333e-05,
|
| 2185 |
+
"loss": 0.335,
|
| 2186 |
+
"step": 1510
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 7.444717444717445,
|
| 2190 |
+
"grad_norm": 0.84765625,
|
| 2191 |
+
"learning_rate": 3.671448787054571e-05,
|
| 2192 |
+
"loss": 0.3305,
|
| 2193 |
+
"step": 1515
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.469287469287469,
|
| 2197 |
+
"grad_norm": 0.8515625,
|
| 2198 |
+
"learning_rate": 3.605114224225028e-05,
|
| 2199 |
+
"loss": 0.332,
|
| 2200 |
+
"step": 1520
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 7.493857493857494,
|
| 2204 |
+
"grad_norm": 0.89453125,
|
| 2205 |
+
"learning_rate": 3.5392523692033006e-05,
|
| 2206 |
+
"loss": 0.3261,
|
| 2207 |
+
"step": 1525
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.518427518427519,
|
| 2211 |
+
"grad_norm": 1.09375,
|
| 2212 |
+
"learning_rate": 3.473868090475574e-05,
|
| 2213 |
+
"loss": 0.3363,
|
| 2214 |
+
"step": 1530
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 7.542997542997543,
|
| 2218 |
+
"grad_norm": 0.88671875,
|
| 2219 |
+
"learning_rate": 3.408966221225773e-05,
|
| 2220 |
+
"loss": 0.3239,
|
| 2221 |
+
"step": 1535
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 7.5675675675675675,
|
| 2225 |
+
"grad_norm": 0.875,
|
| 2226 |
+
"learning_rate": 3.3445515589782574e-05,
|
| 2227 |
+
"loss": 0.3301,
|
| 2228 |
+
"step": 1540
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.592137592137592,
|
| 2232 |
+
"grad_norm": 0.875,
|
| 2233 |
+
"learning_rate": 3.2806288652432174e-05,
|
| 2234 |
+
"loss": 0.3228,
|
| 2235 |
+
"step": 1545
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.616707616707616,
|
| 2239 |
+
"grad_norm": 0.8828125,
|
| 2240 |
+
"learning_rate": 3.217202865164697e-05,
|
| 2241 |
+
"loss": 0.3389,
|
| 2242 |
+
"step": 1550
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 7.641277641277641,
|
| 2246 |
+
"grad_norm": 0.859375,
|
| 2247 |
+
"learning_rate": 3.154278247171314e-05,
|
| 2248 |
+
"loss": 0.3302,
|
| 2249 |
+
"step": 1555
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.665847665847666,
|
| 2253 |
+
"grad_norm": 0.89453125,
|
| 2254 |
+
"learning_rate": 3.09185966262968e-05,
|
| 2255 |
+
"loss": 0.3245,
|
| 2256 |
+
"step": 1560
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 7.69041769041769,
|
| 2260 |
+
"grad_norm": 0.86328125,
|
| 2261 |
+
"learning_rate": 3.0299517255005937e-05,
|
| 2262 |
+
"loss": 0.3306,
|
| 2263 |
+
"step": 1565
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 7.714987714987715,
|
| 2267 |
+
"grad_norm": 0.90625,
|
| 2268 |
+
"learning_rate": 2.9685590119979688e-05,
|
| 2269 |
+
"loss": 0.3322,
|
| 2270 |
+
"step": 1570
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 7.739557739557739,
|
| 2274 |
+
"grad_norm": 0.8828125,
|
| 2275 |
+
"learning_rate": 2.9076860602505564e-05,
|
| 2276 |
+
"loss": 0.3327,
|
| 2277 |
+
"step": 1575
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 7.764127764127764,
|
| 2281 |
+
"grad_norm": 0.953125,
|
| 2282 |
+
"learning_rate": 2.8473373699664997e-05,
|
| 2283 |
+
"loss": 0.3342,
|
| 2284 |
+
"step": 1580
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 7.788697788697789,
|
| 2288 |
+
"grad_norm": 0.89453125,
|
| 2289 |
+
"learning_rate": 2.7875174021007e-05,
|
| 2290 |
+
"loss": 0.3318,
|
| 2291 |
+
"step": 1585
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 7.813267813267814,
|
| 2295 |
+
"grad_norm": 0.85546875,
|
| 2296 |
+
"learning_rate": 2.728230578525086e-05,
|
| 2297 |
+
"loss": 0.3357,
|
| 2298 |
+
"step": 1590
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 7.837837837837838,
|
| 2302 |
+
"grad_norm": 0.8984375,
|
| 2303 |
+
"learning_rate": 2.669481281701739e-05,
|
| 2304 |
+
"loss": 0.3345,
|
| 2305 |
+
"step": 1595
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 7.862407862407863,
|
| 2309 |
+
"grad_norm": 0.94921875,
|
| 2310 |
+
"learning_rate": 2.6112738543589312e-05,
|
| 2311 |
+
"loss": 0.3324,
|
| 2312 |
+
"step": 1600
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 7.886977886977887,
|
| 2316 |
+
"grad_norm": 0.86328125,
|
| 2317 |
+
"learning_rate": 2.553612599170143e-05,
|
| 2318 |
+
"loss": 0.3278,
|
| 2319 |
+
"step": 1605
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 7.9115479115479115,
|
| 2323 |
+
"grad_norm": 0.8046875,
|
| 2324 |
+
"learning_rate": 2.496501778435977e-05,
|
| 2325 |
+
"loss": 0.323,
|
| 2326 |
+
"step": 1610
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 7.936117936117936,
|
| 2330 |
+
"grad_norm": 0.953125,
|
| 2331 |
+
"learning_rate": 2.4399456137691147e-05,
|
| 2332 |
+
"loss": 0.3364,
|
| 2333 |
+
"step": 1615
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 7.96068796068796,
|
| 2337 |
+
"grad_norm": 0.85546875,
|
| 2338 |
+
"learning_rate": 2.3839482857822458e-05,
|
| 2339 |
+
"loss": 0.3348,
|
| 2340 |
+
"step": 1620
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 7.985257985257985,
|
| 2344 |
+
"grad_norm": 0.92578125,
|
| 2345 |
+
"learning_rate": 2.328513933779034e-05,
|
| 2346 |
+
"loss": 0.3321,
|
| 2347 |
+
"step": 1625
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 8.0,
|
| 2351 |
+
"eval_loss": 4.43484354019165,
|
| 2352 |
+
"eval_runtime": 2.0469,
|
| 2353 |
+
"eval_samples_per_second": 4.885,
|
| 2354 |
+
"eval_steps_per_second": 0.977,
|
| 2355 |
+
"step": 1628
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 8.00982800982801,
|
| 2359 |
+
"grad_norm": 0.703125,
|
| 2360 |
+
"learning_rate": 2.2736466554481617e-05,
|
| 2361 |
+
"loss": 0.3114,
|
| 2362 |
+
"step": 1630
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 8.034398034398034,
|
| 2366 |
+
"grad_norm": 0.78125,
|
| 2367 |
+
"learning_rate": 2.2193505065604014e-05,
|
| 2368 |
+
"loss": 0.3037,
|
| 2369 |
+
"step": 1635
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 8.058968058968059,
|
| 2373 |
+
"grad_norm": 0.78515625,
|
| 2374 |
+
"learning_rate": 2.1656295006688353e-05,
|
| 2375 |
+
"loss": 0.2959,
|
| 2376 |
+
"step": 1640
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 8.083538083538084,
|
| 2380 |
+
"grad_norm": 0.88671875,
|
| 2381 |
+
"learning_rate": 2.1124876088121692e-05,
|
| 2382 |
+
"loss": 0.3102,
|
| 2383 |
+
"step": 1645
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 8.108108108108109,
|
| 2387 |
+
"grad_norm": 0.83984375,
|
| 2388 |
+
"learning_rate": 2.0599287592211968e-05,
|
| 2389 |
+
"loss": 0.2975,
|
| 2390 |
+
"step": 1650
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 8.132678132678132,
|
| 2394 |
+
"grad_norm": 0.82421875,
|
| 2395 |
+
"learning_rate": 2.0079568370284128e-05,
|
| 2396 |
+
"loss": 0.2934,
|
| 2397 |
+
"step": 1655
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 8.157248157248157,
|
| 2401 |
+
"grad_norm": 0.84765625,
|
| 2402 |
+
"learning_rate": 1.956575683980846e-05,
|
| 2403 |
+
"loss": 0.299,
|
| 2404 |
+
"step": 1660
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 8.181818181818182,
|
| 2408 |
+
"grad_norm": 0.81640625,
|
| 2409 |
+
"learning_rate": 1.9057890981560677e-05,
|
| 2410 |
+
"loss": 0.2952,
|
| 2411 |
+
"step": 1665
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 8.206388206388207,
|
| 2415 |
+
"grad_norm": 0.8046875,
|
| 2416 |
+
"learning_rate": 1.85560083368143e-05,
|
| 2417 |
+
"loss": 0.3067,
|
| 2418 |
+
"step": 1670
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 8.230958230958231,
|
| 2422 |
+
"grad_norm": 0.8203125,
|
| 2423 |
+
"learning_rate": 1.806014600456588e-05,
|
| 2424 |
+
"loss": 0.2978,
|
| 2425 |
+
"step": 1675
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 8.255528255528256,
|
| 2429 |
+
"grad_norm": 0.80078125,
|
| 2430 |
+
"learning_rate": 1.757034063879235e-05,
|
| 2431 |
+
"loss": 0.2973,
|
| 2432 |
+
"step": 1680
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 8.28009828009828,
|
| 2436 |
+
"grad_norm": 0.796875,
|
| 2437 |
+
"learning_rate": 1.708662844574178e-05,
|
| 2438 |
+
"loss": 0.2983,
|
| 2439 |
+
"step": 1685
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 8.304668304668304,
|
| 2443 |
+
"grad_norm": 0.78125,
|
| 2444 |
+
"learning_rate": 1.6609045181256976e-05,
|
| 2445 |
+
"loss": 0.2991,
|
| 2446 |
+
"step": 1690
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 8.32923832923833,
|
| 2450 |
+
"grad_norm": 0.8203125,
|
| 2451 |
+
"learning_rate": 1.61376261481323e-05,
|
| 2452 |
+
"loss": 0.3102,
|
| 2453 |
+
"step": 1695
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 8.353808353808354,
|
| 2457 |
+
"grad_norm": 0.8828125,
|
| 2458 |
+
"learning_rate": 1.5672406193504384e-05,
|
| 2459 |
+
"loss": 0.2984,
|
| 2460 |
+
"step": 1700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 8.378378378378379,
|
| 2464 |
+
"grad_norm": 0.7578125,
|
| 2465 |
+
"learning_rate": 1.5213419706275878e-05,
|
| 2466 |
+
"loss": 0.298,
|
| 2467 |
+
"step": 1705
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 8.402948402948404,
|
| 2471 |
+
"grad_norm": 0.8046875,
|
| 2472 |
+
"learning_rate": 1.4760700614573731e-05,
|
| 2473 |
+
"loss": 0.3021,
|
| 2474 |
+
"step": 1710
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 8.427518427518427,
|
| 2478 |
+
"grad_norm": 0.84375,
|
| 2479 |
+
"learning_rate": 1.4314282383241096e-05,
|
| 2480 |
+
"loss": 0.3068,
|
| 2481 |
+
"step": 1715
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 8.452088452088452,
|
| 2485 |
+
"grad_norm": 0.82421875,
|
| 2486 |
+
"learning_rate": 1.3874198011363582e-05,
|
| 2487 |
+
"loss": 0.3038,
|
| 2488 |
+
"step": 1720
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 8.476658476658477,
|
| 2492 |
+
"grad_norm": 0.8359375,
|
| 2493 |
+
"learning_rate": 1.3440480029830127e-05,
|
| 2494 |
+
"loss": 0.3024,
|
| 2495 |
+
"step": 1725
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 8.501228501228502,
|
| 2499 |
+
"grad_norm": 0.86328125,
|
| 2500 |
+
"learning_rate": 1.301316049892818e-05,
|
| 2501 |
+
"loss": 0.3018,
|
| 2502 |
+
"step": 1730
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 8.525798525798526,
|
| 2506 |
+
"grad_norm": 0.87890625,
|
| 2507 |
+
"learning_rate": 1.2592271005973888e-05,
|
| 2508 |
+
"loss": 0.3034,
|
| 2509 |
+
"step": 1735
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 8.55036855036855,
|
| 2513 |
+
"grad_norm": 0.8046875,
|
| 2514 |
+
"learning_rate": 1.2177842662977135e-05,
|
| 2515 |
+
"loss": 0.2999,
|
| 2516 |
+
"step": 1740
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 8.574938574938574,
|
| 2520 |
+
"grad_norm": 0.91796875,
|
| 2521 |
+
"learning_rate": 1.1769906104341832e-05,
|
| 2522 |
+
"loss": 0.2997,
|
| 2523 |
+
"step": 1745
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 8.5995085995086,
|
| 2527 |
+
"grad_norm": 0.8203125,
|
| 2528 |
+
"learning_rate": 1.136849148460125e-05,
|
| 2529 |
+
"loss": 0.3042,
|
| 2530 |
+
"step": 1750
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 8.624078624078624,
|
| 2534 |
+
"grad_norm": 0.88671875,
|
| 2535 |
+
"learning_rate": 1.0973628476189257e-05,
|
| 2536 |
+
"loss": 0.2933,
|
| 2537 |
+
"step": 1755
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 8.64864864864865,
|
| 2541 |
+
"grad_norm": 0.8046875,
|
| 2542 |
+
"learning_rate": 1.0585346267246743e-05,
|
| 2543 |
+
"loss": 0.2983,
|
| 2544 |
+
"step": 1760
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 8.673218673218674,
|
| 2548 |
+
"grad_norm": 0.7890625,
|
| 2549 |
+
"learning_rate": 1.0203673559464089e-05,
|
| 2550 |
+
"loss": 0.2963,
|
| 2551 |
+
"step": 1765
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 8.697788697788697,
|
| 2555 |
+
"grad_norm": 0.83984375,
|
| 2556 |
+
"learning_rate": 9.82863856595968e-06,
|
| 2557 |
+
"loss": 0.2952,
|
| 2558 |
+
"step": 1770
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 8.722358722358722,
|
| 2562 |
+
"grad_norm": 0.8359375,
|
| 2563 |
+
"learning_rate": 9.460269009194167e-06,
|
| 2564 |
+
"loss": 0.304,
|
| 2565 |
+
"step": 1775
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 8.746928746928747,
|
| 2569 |
+
"grad_norm": 0.84765625,
|
| 2570 |
+
"learning_rate": 9.098592118921435e-06,
|
| 2571 |
+
"loss": 0.3037,
|
| 2572 |
+
"step": 1780
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 8.771498771498772,
|
| 2576 |
+
"grad_norm": 0.79296875,
|
| 2577 |
+
"learning_rate": 8.74363463017569e-06,
|
| 2578 |
+
"loss": 0.3058,
|
| 2579 |
+
"step": 1785
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 8.796068796068797,
|
| 2583 |
+
"grad_norm": 1.03125,
|
| 2584 |
+
"learning_rate": 8.395422781295192e-06,
|
| 2585 |
+
"loss": 0.3031,
|
| 2586 |
+
"step": 1790
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 8.82063882063882,
|
| 2590 |
+
"grad_norm": 0.875,
|
| 2591 |
+
"learning_rate": 8.053982311982867e-06,
|
| 2592 |
+
"loss": 0.3017,
|
| 2593 |
+
"step": 1795
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 8.845208845208845,
|
| 2597 |
+
"grad_norm": 0.9453125,
|
| 2598 |
+
"learning_rate": 7.719338461403435e-06,
|
| 2599 |
+
"loss": 0.2978,
|
| 2600 |
+
"step": 1800
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 8.86977886977887,
|
| 2604 |
+
"grad_norm": 0.82421875,
|
| 2605 |
+
"learning_rate": 7.3915159663179075e-06,
|
| 2606 |
+
"loss": 0.2905,
|
| 2607 |
+
"step": 1805
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 8.894348894348894,
|
| 2611 |
+
"grad_norm": 0.81640625,
|
| 2612 |
+
"learning_rate": 7.070539059254977e-06,
|
| 2613 |
+
"loss": 0.2957,
|
| 2614 |
+
"step": 1810
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 8.91891891891892,
|
| 2618 |
+
"grad_norm": 0.7890625,
|
| 2619 |
+
"learning_rate": 6.756431466719737e-06,
|
| 2620 |
+
"loss": 0.2988,
|
| 2621 |
+
"step": 1815
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 8.943488943488944,
|
| 2625 |
+
"grad_norm": 0.84375,
|
| 2626 |
+
"learning_rate": 6.4492164074399065e-06,
|
| 2627 |
+
"loss": 0.3019,
|
| 2628 |
+
"step": 1820
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 8.968058968058967,
|
| 2632 |
+
"grad_norm": 0.8359375,
|
| 2633 |
+
"learning_rate": 6.148916590649434e-06,
|
| 2634 |
+
"loss": 0.3051,
|
| 2635 |
+
"step": 1825
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 8.992628992628992,
|
| 2639 |
+
"grad_norm": 0.859375,
|
| 2640 |
+
"learning_rate": 5.8555542144098865e-06,
|
| 2641 |
+
"loss": 0.3089,
|
| 2642 |
+
"step": 1830
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 8.997542997542997,
|
| 2646 |
+
"eval_loss": 4.689078330993652,
|
| 2647 |
+
"eval_runtime": 2.054,
|
| 2648 |
+
"eval_samples_per_second": 4.868,
|
| 2649 |
+
"eval_steps_per_second": 0.974,
|
| 2650 |
+
"step": 1831
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 9.017199017199017,
|
| 2654 |
+
"grad_norm": 0.78515625,
|
| 2655 |
+
"learning_rate": 5.569150963969494e-06,
|
| 2656 |
+
"loss": 0.2973,
|
| 2657 |
+
"step": 1835
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 9.041769041769042,
|
| 2661 |
+
"grad_norm": 0.80078125,
|
| 2662 |
+
"learning_rate": 5.289728010160366e-06,
|
| 2663 |
+
"loss": 0.3,
|
| 2664 |
+
"step": 1840
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 9.066339066339067,
|
| 2668 |
+
"grad_norm": 0.7578125,
|
| 2669 |
+
"learning_rate": 5.0173060078333225e-06,
|
| 2670 |
+
"loss": 0.299,
|
| 2671 |
+
"step": 1845
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 9.090909090909092,
|
| 2675 |
+
"grad_norm": 0.796875,
|
| 2676 |
+
"learning_rate": 4.7519050943312325e-06,
|
| 2677 |
+
"loss": 0.2971,
|
| 2678 |
+
"step": 1850
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 9.115479115479115,
|
| 2682 |
+
"grad_norm": 0.78125,
|
| 2683 |
+
"learning_rate": 4.493544888000467e-06,
|
| 2684 |
+
"loss": 0.2968,
|
| 2685 |
+
"step": 1855
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 9.14004914004914,
|
| 2689 |
+
"grad_norm": 0.765625,
|
| 2690 |
+
"learning_rate": 4.242244486740643e-06,
|
| 2691 |
+
"loss": 0.295,
|
| 2692 |
+
"step": 1860
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 9.164619164619165,
|
| 2696 |
+
"grad_norm": 0.8125,
|
| 2697 |
+
"learning_rate": 3.99802246659301e-06,
|
| 2698 |
+
"loss": 0.2911,
|
| 2699 |
+
"step": 1865
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 9.18918918918919,
|
| 2703 |
+
"grad_norm": 0.76953125,
|
| 2704 |
+
"learning_rate": 3.760896880367215e-06,
|
| 2705 |
+
"loss": 0.2916,
|
| 2706 |
+
"step": 1870
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 9.213759213759214,
|
| 2710 |
+
"grad_norm": 0.7890625,
|
| 2711 |
+
"learning_rate": 3.530885256306915e-06,
|
| 2712 |
+
"loss": 0.2998,
|
| 2713 |
+
"step": 1875
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 9.238329238329237,
|
| 2717 |
+
"grad_norm": 0.8046875,
|
| 2718 |
+
"learning_rate": 3.308004596794101e-06,
|
| 2719 |
+
"loss": 0.2948,
|
| 2720 |
+
"step": 1880
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 9.262899262899262,
|
| 2724 |
+
"grad_norm": 0.81640625,
|
| 2725 |
+
"learning_rate": 3.092271377092215e-06,
|
| 2726 |
+
"loss": 0.2917,
|
| 2727 |
+
"step": 1885
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 9.287469287469287,
|
| 2731 |
+
"grad_norm": 0.75,
|
| 2732 |
+
"learning_rate": 2.8837015441283586e-06,
|
| 2733 |
+
"loss": 0.295,
|
| 2734 |
+
"step": 1890
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 9.312039312039312,
|
| 2738 |
+
"grad_norm": 0.74609375,
|
| 2739 |
+
"learning_rate": 2.682310515314512e-06,
|
| 2740 |
+
"loss": 0.2928,
|
| 2741 |
+
"step": 1895
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 9.336609336609337,
|
| 2745 |
+
"grad_norm": 0.8046875,
|
| 2746 |
+
"learning_rate": 2.488113177407869e-06,
|
| 2747 |
+
"loss": 0.2909,
|
| 2748 |
+
"step": 1900
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 9.361179361179362,
|
| 2752 |
+
"grad_norm": 0.75,
|
| 2753 |
+
"learning_rate": 2.3011238854103947e-06,
|
| 2754 |
+
"loss": 0.2929,
|
| 2755 |
+
"step": 1905
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 9.385749385749385,
|
| 2759 |
+
"grad_norm": 0.796875,
|
| 2760 |
+
"learning_rate": 2.1213564615077065e-06,
|
| 2761 |
+
"loss": 0.2936,
|
| 2762 |
+
"step": 1910
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 9.41031941031941,
|
| 2766 |
+
"grad_norm": 0.79296875,
|
| 2767 |
+
"learning_rate": 1.9488241940473828e-06,
|
| 2768 |
+
"loss": 0.2941,
|
| 2769 |
+
"step": 1915
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 9.434889434889435,
|
| 2773 |
+
"grad_norm": 0.76171875,
|
| 2774 |
+
"learning_rate": 1.783539836556669e-06,
|
| 2775 |
+
"loss": 0.2941,
|
| 2776 |
+
"step": 1920
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 9.45945945945946,
|
| 2780 |
+
"grad_norm": 0.78515625,
|
| 2781 |
+
"learning_rate": 1.6255156067997323e-06,
|
| 2782 |
+
"loss": 0.2869,
|
| 2783 |
+
"step": 1925
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 9.484029484029485,
|
| 2787 |
+
"grad_norm": 0.80859375,
|
| 2788 |
+
"learning_rate": 1.474763185874517e-06,
|
| 2789 |
+
"loss": 0.3029,
|
| 2790 |
+
"step": 1930
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 9.50859950859951,
|
| 2794 |
+
"grad_norm": 0.80859375,
|
| 2795 |
+
"learning_rate": 1.3312937173493577e-06,
|
| 2796 |
+
"loss": 0.2943,
|
| 2797 |
+
"step": 1935
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 9.533169533169533,
|
| 2801 |
+
"grad_norm": 0.83984375,
|
| 2802 |
+
"learning_rate": 1.19511780643915e-06,
|
| 2803 |
+
"loss": 0.2963,
|
| 2804 |
+
"step": 1940
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 9.557739557739557,
|
| 2808 |
+
"grad_norm": 0.765625,
|
| 2809 |
+
"learning_rate": 1.066245519221465e-06,
|
| 2810 |
+
"loss": 0.3002,
|
| 2811 |
+
"step": 1945
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 9.582309582309582,
|
| 2815 |
+
"grad_norm": 0.76171875,
|
| 2816 |
+
"learning_rate": 9.446863818924679e-07,
|
| 2817 |
+
"loss": 0.2988,
|
| 2818 |
+
"step": 1950
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 9.606879606879607,
|
| 2822 |
+
"grad_norm": 0.77734375,
|
| 2823 |
+
"learning_rate": 8.304493800627589e-07,
|
| 2824 |
+
"loss": 0.297,
|
| 2825 |
+
"step": 1955
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 9.631449631449632,
|
| 2829 |
+
"grad_norm": 0.75,
|
| 2830 |
+
"learning_rate": 7.235429580931152e-07,
|
| 2831 |
+
"loss": 0.2902,
|
| 2832 |
+
"step": 1960
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 9.656019656019655,
|
| 2836 |
+
"grad_norm": 0.859375,
|
| 2837 |
+
"learning_rate": 6.239750184703464e-07,
|
| 2838 |
+
"loss": 0.2933,
|
| 2839 |
+
"step": 1965
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 9.68058968058968,
|
| 2843 |
+
"grad_norm": 0.76953125,
|
| 2844 |
+
"learning_rate": 5.317529212230721e-07,
|
| 2845 |
+
"loss": 0.2943,
|
| 2846 |
+
"step": 1970
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 9.705159705159705,
|
| 2850 |
+
"grad_norm": 0.78125,
|
| 2851 |
+
"learning_rate": 4.4688348337774686e-07,
|
| 2852 |
+
"loss": 0.3049,
|
| 2853 |
+
"step": 1975
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 9.72972972972973,
|
| 2857 |
+
"grad_norm": 0.83984375,
|
| 2858 |
+
"learning_rate": 3.693729784546962e-07,
|
| 2859 |
+
"loss": 0.2876,
|
| 2860 |
+
"step": 1980
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 9.754299754299755,
|
| 2864 |
+
"grad_norm": 0.78515625,
|
| 2865 |
+
"learning_rate": 2.9922713600439854e-07,
|
| 2866 |
+
"loss": 0.2971,
|
| 2867 |
+
"step": 1985
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 9.77886977886978,
|
| 2871 |
+
"grad_norm": 0.765625,
|
| 2872 |
+
"learning_rate": 2.3645114118395762e-07,
|
| 2873 |
+
"loss": 0.288,
|
| 2874 |
+
"step": 1990
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 9.803439803439803,
|
| 2878 |
+
"grad_norm": 0.84765625,
|
| 2879 |
+
"learning_rate": 1.8104963437381993e-07,
|
| 2880 |
+
"loss": 0.2951,
|
| 2881 |
+
"step": 1995
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 9.828009828009828,
|
| 2885 |
+
"grad_norm": 0.7578125,
|
| 2886 |
+
"learning_rate": 1.3302671083474938e-07,
|
| 2887 |
+
"loss": 0.2927,
|
| 2888 |
+
"step": 2000
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 9.852579852579852,
|
| 2892 |
+
"grad_norm": 0.78515625,
|
| 2893 |
+
"learning_rate": 9.238592040512472e-08,
|
| 2894 |
+
"loss": 0.2932,
|
| 2895 |
+
"step": 2005
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 9.877149877149877,
|
| 2899 |
+
"grad_norm": 0.75390625,
|
| 2900 |
+
"learning_rate": 5.913026723850523e-08,
|
| 2901 |
+
"loss": 0.2913,
|
| 2902 |
+
"step": 2010
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 9.901719901719902,
|
| 2906 |
+
"grad_norm": 0.8046875,
|
| 2907 |
+
"learning_rate": 3.3262209581619297e-08,
|
| 2908 |
+
"loss": 0.2997,
|
| 2909 |
+
"step": 2015
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 9.926289926289925,
|
| 2913 |
+
"grad_norm": 0.85546875,
|
| 2914 |
+
"learning_rate": 1.4783659592576548e-08,
|
| 2915 |
+
"loss": 0.2977,
|
| 2916 |
+
"step": 2020
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 9.95085995085995,
|
| 2920 |
+
"grad_norm": 0.80078125,
|
| 2921 |
+
"learning_rate": 3.6959831996030704e-09,
|
| 2922 |
+
"loss": 0.2988,
|
| 2923 |
+
"step": 2025
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 9.975429975429975,
|
| 2927 |
+
"grad_norm": 0.84375,
|
| 2928 |
+
"learning_rate": 0.0,
|
| 2929 |
+
"loss": 0.3016,
|
| 2930 |
+
"step": 2030
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 9.975429975429975,
|
| 2934 |
+
"eval_loss": 4.717045307159424,
|
| 2935 |
+
"eval_runtime": 2.0451,
|
| 2936 |
+
"eval_samples_per_second": 4.89,
|
| 2937 |
+
"eval_steps_per_second": 0.978,
|
| 2938 |
+
"step": 2030
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 9.975429975429975,
|
| 2942 |
+
"step": 2030,
|
| 2943 |
+
"total_flos": 1.5518062706111283e+18,
|
| 2944 |
+
"train_loss": 1.3545548074938394,
|
| 2945 |
+
"train_runtime": 13205.4864,
|
| 2946 |
+
"train_samples_per_second": 2.465,
|
| 2947 |
+
"train_steps_per_second": 0.154
|
| 2948 |
+
}
|
| 2949 |
+
],
|
| 2950 |
+
"logging_steps": 5,
|
| 2951 |
+
"max_steps": 2030,
|
| 2952 |
+
"num_input_tokens_seen": 0,
|
| 2953 |
+
"num_train_epochs": 10,
|
| 2954 |
+
"save_steps": 100,
|
| 2955 |
+
"total_flos": 1.5518062706111283e+18,
|
| 2956 |
+
"train_batch_size": 4,
|
| 2957 |
+
"trial_name": null,
|
| 2958 |
+
"trial_params": null
|
| 2959 |
+
}
|