Model save
Browse files- README.md +78 -0
- adapter_model.safetensors +1 -1
- all_results.json +9 -0
- runs/Jun09_15-53-33_e681ffd796e5/events.out.tfevents.1717948893.e681ffd796e5.16736.0 +2 -2
- train_results.json +9 -0
- trainer_state.json +3013 -0
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-1.1-2b-it
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: gemma2b-it-1.1-summarize-gpt4o-256k
|
| 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 |
+
# gemma2b-it-1.1-summarize-gpt4o-256k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [google/gemma-1.1-2b-it](https://huggingface.co/google/gemma-1.1-2b-it) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.7127
|
| 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: 8
|
| 44 |
+
- eval_batch_size: 4
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 8
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 128
|
| 50 |
+
- total_eval_batch_size: 32
|
| 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 |
+
| 0.9633 | 0.9976 | 206 | 2.6959 |
|
| 61 |
+
| 0.865 | 2.0 | 413 | 2.6511 |
|
| 62 |
+
| 0.8266 | 2.9976 | 619 | 2.6475 |
|
| 63 |
+
| 0.7953 | 4.0 | 826 | 2.6603 |
|
| 64 |
+
| 0.7708 | 4.9976 | 1032 | 2.6720 |
|
| 65 |
+
| 0.75 | 6.0 | 1239 | 2.6898 |
|
| 66 |
+
| 0.7446 | 6.9976 | 1445 | 2.7026 |
|
| 67 |
+
| 0.7301 | 8.0 | 1652 | 2.7081 |
|
| 68 |
+
| 0.7268 | 8.9976 | 1858 | 2.7128 |
|
| 69 |
+
| 0.7318 | 9.9758 | 2060 | 2.7127 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- PEFT 0.11.1
|
| 75 |
+
- Transformers 4.41.2
|
| 76 |
+
- Pytorch 2.3.1+cu121
|
| 77 |
+
- Datasets 2.19.2
|
| 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 39256960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b947202dd42338b22fba341715e6fa49bfc9a606a5f06c59b558c0d89a8b2d49
|
| 3 |
size 39256960
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.975786924939467,
|
| 3 |
+
"total_flos": 3.242523538363515e+18,
|
| 4 |
+
"train_loss": 0.8949183881861492,
|
| 5 |
+
"train_runtime": 11275.6837,
|
| 6 |
+
"train_samples": 256423,
|
| 7 |
+
"train_samples_per_second": 23.387,
|
| 8 |
+
"train_steps_per_second": 0.183
|
| 9 |
+
}
|
runs/Jun09_15-53-33_e681ffd796e5/events.out.tfevents.1717948893.e681ffd796e5.16736.0
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 sha256:5b0b837ecb0591ec3e5c72e8d12e5de15dccbcde57e8da00dd4a860d5664a77d
|
| 3 |
+
size 95714
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.975786924939467,
|
| 3 |
+
"total_flos": 3.242523538363515e+18,
|
| 4 |
+
"train_loss": 0.8949183881861492,
|
| 5 |
+
"train_runtime": 11275.6837,
|
| 6 |
+
"train_samples": 256423,
|
| 7 |
+
"train_samples_per_second": 23.387,
|
| 8 |
+
"train_steps_per_second": 0.183
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3013 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.975786924939467,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2060,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004842615012106538,
|
| 13 |
+
"grad_norm": 3.9375,
|
| 14 |
+
"learning_rate": 9.70873786407767e-07,
|
| 15 |
+
"loss": 5.486,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.024213075060532687,
|
| 20 |
+
"grad_norm": 3.875,
|
| 21 |
+
"learning_rate": 4.854368932038835e-06,
|
| 22 |
+
"loss": 5.4688,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.048426150121065374,
|
| 27 |
+
"grad_norm": 3.9375,
|
| 28 |
+
"learning_rate": 9.70873786407767e-06,
|
| 29 |
+
"loss": 5.4581,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07263922518159806,
|
| 34 |
+
"grad_norm": 3.796875,
|
| 35 |
+
"learning_rate": 1.4563106796116505e-05,
|
| 36 |
+
"loss": 5.3204,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09685230024213075,
|
| 41 |
+
"grad_norm": 3.578125,
|
| 42 |
+
"learning_rate": 1.941747572815534e-05,
|
| 43 |
+
"loss": 5.0021,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12106537530266344,
|
| 48 |
+
"grad_norm": 3.46875,
|
| 49 |
+
"learning_rate": 2.4271844660194176e-05,
|
| 50 |
+
"loss": 4.5571,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.14527845036319612,
|
| 55 |
+
"grad_norm": 3.265625,
|
| 56 |
+
"learning_rate": 2.912621359223301e-05,
|
| 57 |
+
"loss": 3.9862,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1694915254237288,
|
| 62 |
+
"grad_norm": 2.84375,
|
| 63 |
+
"learning_rate": 3.398058252427185e-05,
|
| 64 |
+
"loss": 3.3741,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1937046004842615,
|
| 69 |
+
"grad_norm": 2.34375,
|
| 70 |
+
"learning_rate": 3.883495145631068e-05,
|
| 71 |
+
"loss": 3.0035,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2179176755447942,
|
| 76 |
+
"grad_norm": 2.8125,
|
| 77 |
+
"learning_rate": 4.368932038834951e-05,
|
| 78 |
+
"loss": 2.7678,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.24213075060532688,
|
| 83 |
+
"grad_norm": 3.046875,
|
| 84 |
+
"learning_rate": 4.854368932038835e-05,
|
| 85 |
+
"loss": 2.4982,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26634382566585957,
|
| 90 |
+
"grad_norm": 3.28125,
|
| 91 |
+
"learning_rate": 5.339805825242719e-05,
|
| 92 |
+
"loss": 2.2149,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.29055690072639223,
|
| 97 |
+
"grad_norm": 8.625,
|
| 98 |
+
"learning_rate": 5.825242718446602e-05,
|
| 99 |
+
"loss": 1.9807,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.31476997578692495,
|
| 104 |
+
"grad_norm": 8.875,
|
| 105 |
+
"learning_rate": 6.310679611650486e-05,
|
| 106 |
+
"loss": 1.9918,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3389830508474576,
|
| 111 |
+
"grad_norm": 2.234375,
|
| 112 |
+
"learning_rate": 6.79611650485437e-05,
|
| 113 |
+
"loss": 1.8154,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.36319612590799033,
|
| 118 |
+
"grad_norm": 2.671875,
|
| 119 |
+
"learning_rate": 7.281553398058253e-05,
|
| 120 |
+
"loss": 1.7162,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.387409200968523,
|
| 125 |
+
"grad_norm": 1.671875,
|
| 126 |
+
"learning_rate": 7.766990291262136e-05,
|
| 127 |
+
"loss": 1.6152,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4116222760290557,
|
| 132 |
+
"grad_norm": 0.63671875,
|
| 133 |
+
"learning_rate": 8.25242718446602e-05,
|
| 134 |
+
"loss": 1.5063,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4358353510895884,
|
| 139 |
+
"grad_norm": 0.69921875,
|
| 140 |
+
"learning_rate": 8.737864077669902e-05,
|
| 141 |
+
"loss": 1.4152,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4600484261501211,
|
| 146 |
+
"grad_norm": 0.73046875,
|
| 147 |
+
"learning_rate": 9.223300970873788e-05,
|
| 148 |
+
"loss": 1.3459,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.48426150121065376,
|
| 153 |
+
"grad_norm": 0.41015625,
|
| 154 |
+
"learning_rate": 9.70873786407767e-05,
|
| 155 |
+
"loss": 1.3005,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5084745762711864,
|
| 160 |
+
"grad_norm": 0.8125,
|
| 161 |
+
"learning_rate": 0.00010194174757281553,
|
| 162 |
+
"loss": 1.2498,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5326876513317191,
|
| 167 |
+
"grad_norm": 0.5234375,
|
| 168 |
+
"learning_rate": 0.00010679611650485437,
|
| 169 |
+
"loss": 1.218,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5569007263922519,
|
| 174 |
+
"grad_norm": 0.4453125,
|
| 175 |
+
"learning_rate": 0.0001116504854368932,
|
| 176 |
+
"loss": 1.1886,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5811138014527845,
|
| 181 |
+
"grad_norm": 0.279296875,
|
| 182 |
+
"learning_rate": 0.00011650485436893204,
|
| 183 |
+
"loss": 1.1448,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6053268765133172,
|
| 188 |
+
"grad_norm": 0.6328125,
|
| 189 |
+
"learning_rate": 0.00012135922330097087,
|
| 190 |
+
"loss": 1.1324,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6295399515738499,
|
| 195 |
+
"grad_norm": 0.443359375,
|
| 196 |
+
"learning_rate": 0.00012621359223300972,
|
| 197 |
+
"loss": 1.1027,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6537530266343826,
|
| 202 |
+
"grad_norm": 0.65234375,
|
| 203 |
+
"learning_rate": 0.00013106796116504855,
|
| 204 |
+
"loss": 1.081,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6779661016949152,
|
| 209 |
+
"grad_norm": 0.412109375,
|
| 210 |
+
"learning_rate": 0.0001359223300970874,
|
| 211 |
+
"loss": 1.0642,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7021791767554479,
|
| 216 |
+
"grad_norm": 0.5078125,
|
| 217 |
+
"learning_rate": 0.00014077669902912623,
|
| 218 |
+
"loss": 1.0491,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7263922518159807,
|
| 223 |
+
"grad_norm": 0.423828125,
|
| 224 |
+
"learning_rate": 0.00014563106796116506,
|
| 225 |
+
"loss": 1.04,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7506053268765133,
|
| 230 |
+
"grad_norm": 0.26953125,
|
| 231 |
+
"learning_rate": 0.00015048543689320389,
|
| 232 |
+
"loss": 1.027,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.774818401937046,
|
| 237 |
+
"grad_norm": 0.447265625,
|
| 238 |
+
"learning_rate": 0.0001553398058252427,
|
| 239 |
+
"loss": 1.0164,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7990314769975787,
|
| 244 |
+
"grad_norm": 0.33203125,
|
| 245 |
+
"learning_rate": 0.00016019417475728157,
|
| 246 |
+
"loss": 1.0065,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8232445520581114,
|
| 251 |
+
"grad_norm": 0.30078125,
|
| 252 |
+
"learning_rate": 0.0001650485436893204,
|
| 253 |
+
"loss": 0.996,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.847457627118644,
|
| 258 |
+
"grad_norm": 0.28125,
|
| 259 |
+
"learning_rate": 0.00016990291262135922,
|
| 260 |
+
"loss": 0.9893,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8716707021791767,
|
| 265 |
+
"grad_norm": 0.310546875,
|
| 266 |
+
"learning_rate": 0.00017475728155339805,
|
| 267 |
+
"loss": 0.9829,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8958837772397095,
|
| 272 |
+
"grad_norm": 0.412109375,
|
| 273 |
+
"learning_rate": 0.0001796116504854369,
|
| 274 |
+
"loss": 0.9746,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9200968523002422,
|
| 279 |
+
"grad_norm": 0.279296875,
|
| 280 |
+
"learning_rate": 0.00018446601941747576,
|
| 281 |
+
"loss": 0.9717,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9443099273607748,
|
| 286 |
+
"grad_norm": 0.30859375,
|
| 287 |
+
"learning_rate": 0.00018932038834951458,
|
| 288 |
+
"loss": 0.9658,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9685230024213075,
|
| 293 |
+
"grad_norm": 0.298828125,
|
| 294 |
+
"learning_rate": 0.0001941747572815534,
|
| 295 |
+
"loss": 0.9654,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.9927360774818402,
|
| 300 |
+
"grad_norm": 0.279296875,
|
| 301 |
+
"learning_rate": 0.00019902912621359224,
|
| 302 |
+
"loss": 0.9633,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.9975786924939467,
|
| 307 |
+
"eval_loss": 2.6959354877471924,
|
| 308 |
+
"eval_runtime": 0.2681,
|
| 309 |
+
"eval_samples_per_second": 37.305,
|
| 310 |
+
"eval_steps_per_second": 3.73,
|
| 311 |
+
"step": 206
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.0169491525423728,
|
| 315 |
+
"grad_norm": 0.7265625,
|
| 316 |
+
"learning_rate": 0.0001999977029597241,
|
| 317 |
+
"loss": 0.9554,
|
| 318 |
+
"step": 210
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.0411622276029056,
|
| 322 |
+
"grad_norm": 0.306640625,
|
| 323 |
+
"learning_rate": 0.00019998837141446378,
|
| 324 |
+
"loss": 0.9453,
|
| 325 |
+
"step": 215
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.0653753026634383,
|
| 329 |
+
"grad_norm": 0.267578125,
|
| 330 |
+
"learning_rate": 0.00019997186246852679,
|
| 331 |
+
"loss": 0.9391,
|
| 332 |
+
"step": 220
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.089588377723971,
|
| 336 |
+
"grad_norm": 0.302734375,
|
| 337 |
+
"learning_rate": 0.00019994817730696412,
|
| 338 |
+
"loss": 0.9382,
|
| 339 |
+
"step": 225
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.1138014527845037,
|
| 343 |
+
"grad_norm": 0.228515625,
|
| 344 |
+
"learning_rate": 0.00019991731762995243,
|
| 345 |
+
"loss": 0.9398,
|
| 346 |
+
"step": 230
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.1380145278450362,
|
| 350 |
+
"grad_norm": 0.275390625,
|
| 351 |
+
"learning_rate": 0.000199879285652672,
|
| 352 |
+
"loss": 0.9253,
|
| 353 |
+
"step": 235
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.162227602905569,
|
| 357 |
+
"grad_norm": 0.27734375,
|
| 358 |
+
"learning_rate": 0.00019983408410514765,
|
| 359 |
+
"loss": 0.9284,
|
| 360 |
+
"step": 240
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.1864406779661016,
|
| 364 |
+
"grad_norm": 0.326171875,
|
| 365 |
+
"learning_rate": 0.00019978171623205292,
|
| 366 |
+
"loss": 0.9222,
|
| 367 |
+
"step": 245
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.2106537530266344,
|
| 371 |
+
"grad_norm": 0.265625,
|
| 372 |
+
"learning_rate": 0.00019972218579247697,
|
| 373 |
+
"loss": 0.9177,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.234866828087167,
|
| 378 |
+
"grad_norm": 0.267578125,
|
| 379 |
+
"learning_rate": 0.00019965549705965488,
|
| 380 |
+
"loss": 0.9149,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.2590799031476998,
|
| 385 |
+
"grad_norm": 0.2333984375,
|
| 386 |
+
"learning_rate": 0.00019958165482066094,
|
| 387 |
+
"loss": 0.9102,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.2832929782082325,
|
| 392 |
+
"grad_norm": 0.330078125,
|
| 393 |
+
"learning_rate": 0.00019950066437606485,
|
| 394 |
+
"loss": 0.9108,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.307506053268765,
|
| 399 |
+
"grad_norm": 0.265625,
|
| 400 |
+
"learning_rate": 0.00019941253153955138,
|
| 401 |
+
"loss": 0.9108,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.331719128329298,
|
| 406 |
+
"grad_norm": 0.232421875,
|
| 407 |
+
"learning_rate": 0.00019931726263750302,
|
| 408 |
+
"loss": 0.9057,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.3559322033898304,
|
| 413 |
+
"grad_norm": 0.2470703125,
|
| 414 |
+
"learning_rate": 0.0001992148645085458,
|
| 415 |
+
"loss": 0.904,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.3801452784503632,
|
| 420 |
+
"grad_norm": 0.384765625,
|
| 421 |
+
"learning_rate": 0.0001991053445030585,
|
| 422 |
+
"loss": 0.9011,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.4043583535108959,
|
| 427 |
+
"grad_norm": 0.34765625,
|
| 428 |
+
"learning_rate": 0.0001989887104826449,
|
| 429 |
+
"loss": 0.9005,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.4285714285714286,
|
| 434 |
+
"grad_norm": 0.322265625,
|
| 435 |
+
"learning_rate": 0.00019886497081956955,
|
| 436 |
+
"loss": 0.8899,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.4527845036319613,
|
| 441 |
+
"grad_norm": 0.2734375,
|
| 442 |
+
"learning_rate": 0.00019873413439615675,
|
| 443 |
+
"loss": 0.8849,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.4769975786924938,
|
| 448 |
+
"grad_norm": 0.24609375,
|
| 449 |
+
"learning_rate": 0.000198596210604153,
|
| 450 |
+
"loss": 0.8902,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.5012106537530268,
|
| 455 |
+
"grad_norm": 0.28515625,
|
| 456 |
+
"learning_rate": 0.00019845120934405274,
|
| 457 |
+
"loss": 0.8939,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.5254237288135593,
|
| 462 |
+
"grad_norm": 0.251953125,
|
| 463 |
+
"learning_rate": 0.00019829914102438779,
|
| 464 |
+
"loss": 0.8943,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.549636803874092,
|
| 469 |
+
"grad_norm": 0.4375,
|
| 470 |
+
"learning_rate": 0.00019814001656098001,
|
| 471 |
+
"loss": 0.8892,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.5738498789346247,
|
| 476 |
+
"grad_norm": 0.35546875,
|
| 477 |
+
"learning_rate": 0.0001979738473761581,
|
| 478 |
+
"loss": 0.881,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.5980629539951574,
|
| 483 |
+
"grad_norm": 0.3046875,
|
| 484 |
+
"learning_rate": 0.00019780064539793723,
|
| 485 |
+
"loss": 0.8819,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.6222760290556901,
|
| 490 |
+
"grad_norm": 0.384765625,
|
| 491 |
+
"learning_rate": 0.00019762042305916316,
|
| 492 |
+
"loss": 0.8727,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.6464891041162226,
|
| 497 |
+
"grad_norm": 0.267578125,
|
| 498 |
+
"learning_rate": 0.0001974331932966196,
|
| 499 |
+
"loss": 0.8815,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.6707021791767556,
|
| 504 |
+
"grad_norm": 0.328125,
|
| 505 |
+
"learning_rate": 0.00019723896955009965,
|
| 506 |
+
"loss": 0.8728,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.694915254237288,
|
| 511 |
+
"grad_norm": 0.58984375,
|
| 512 |
+
"learning_rate": 0.00019703776576144105,
|
| 513 |
+
"loss": 0.8793,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.7191283292978208,
|
| 518 |
+
"grad_norm": 0.296875,
|
| 519 |
+
"learning_rate": 0.00019682959637352541,
|
| 520 |
+
"loss": 0.8735,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.7433414043583535,
|
| 525 |
+
"grad_norm": 0.357421875,
|
| 526 |
+
"learning_rate": 0.00019661447632924144,
|
| 527 |
+
"loss": 0.8746,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.7675544794188862,
|
| 532 |
+
"grad_norm": 0.671875,
|
| 533 |
+
"learning_rate": 0.00019639242107041227,
|
| 534 |
+
"loss": 0.8721,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.791767554479419,
|
| 539 |
+
"grad_norm": 0.3984375,
|
| 540 |
+
"learning_rate": 0.00019616344653668711,
|
| 541 |
+
"loss": 0.8697,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.8159806295399514,
|
| 546 |
+
"grad_norm": 0.349609375,
|
| 547 |
+
"learning_rate": 0.00019592756916439705,
|
| 548 |
+
"loss": 0.8705,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.8401937046004844,
|
| 553 |
+
"grad_norm": 0.4140625,
|
| 554 |
+
"learning_rate": 0.00019568480588537512,
|
| 555 |
+
"loss": 0.8676,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.8644067796610169,
|
| 560 |
+
"grad_norm": 0.283203125,
|
| 561 |
+
"learning_rate": 0.00019543517412574097,
|
| 562 |
+
"loss": 0.8736,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.8886198547215496,
|
| 567 |
+
"grad_norm": 0.453125,
|
| 568 |
+
"learning_rate": 0.00019517869180464992,
|
| 569 |
+
"loss": 0.8663,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.9128329297820823,
|
| 574 |
+
"grad_norm": 0.2890625,
|
| 575 |
+
"learning_rate": 0.00019491537733300676,
|
| 576 |
+
"loss": 0.8699,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.937046004842615,
|
| 581 |
+
"grad_norm": 0.28125,
|
| 582 |
+
"learning_rate": 0.00019464524961214415,
|
| 583 |
+
"loss": 0.8657,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.9612590799031477,
|
| 588 |
+
"grad_norm": 0.28515625,
|
| 589 |
+
"learning_rate": 0.00019436832803246574,
|
| 590 |
+
"loss": 0.8621,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.9854721549636802,
|
| 595 |
+
"grad_norm": 0.23046875,
|
| 596 |
+
"learning_rate": 0.0001940846324720544,
|
| 597 |
+
"loss": 0.865,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 2.0,
|
| 602 |
+
"eval_loss": 2.6510560512542725,
|
| 603 |
+
"eval_runtime": 0.237,
|
| 604 |
+
"eval_samples_per_second": 42.194,
|
| 605 |
+
"eval_steps_per_second": 4.219,
|
| 606 |
+
"step": 413
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.009685230024213,
|
| 610 |
+
"grad_norm": 0.27734375,
|
| 611 |
+
"learning_rate": 0.00019379418329524534,
|
| 612 |
+
"loss": 0.857,
|
| 613 |
+
"step": 415
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.0338983050847457,
|
| 617 |
+
"grad_norm": 0.279296875,
|
| 618 |
+
"learning_rate": 0.00019349700135116416,
|
| 619 |
+
"loss": 0.8445,
|
| 620 |
+
"step": 420
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.0581113801452786,
|
| 624 |
+
"grad_norm": 0.37109375,
|
| 625 |
+
"learning_rate": 0.00019319310797223036,
|
| 626 |
+
"loss": 0.8468,
|
| 627 |
+
"step": 425
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.082324455205811,
|
| 631 |
+
"grad_norm": 0.306640625,
|
| 632 |
+
"learning_rate": 0.00019288252497262612,
|
| 633 |
+
"loss": 0.8413,
|
| 634 |
+
"step": 430
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.106537530266344,
|
| 638 |
+
"grad_norm": 0.259765625,
|
| 639 |
+
"learning_rate": 0.00019256527464673025,
|
| 640 |
+
"loss": 0.8442,
|
| 641 |
+
"step": 435
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.1307506053268765,
|
| 645 |
+
"grad_norm": 0.259765625,
|
| 646 |
+
"learning_rate": 0.00019224137976751795,
|
| 647 |
+
"loss": 0.8374,
|
| 648 |
+
"step": 440
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.154963680387409,
|
| 652 |
+
"grad_norm": 0.3125,
|
| 653 |
+
"learning_rate": 0.0001919108635849261,
|
| 654 |
+
"loss": 0.8399,
|
| 655 |
+
"step": 445
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.179176755447942,
|
| 659 |
+
"grad_norm": 0.306640625,
|
| 660 |
+
"learning_rate": 0.00019157374982418435,
|
| 661 |
+
"loss": 0.837,
|
| 662 |
+
"step": 450
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.2033898305084745,
|
| 666 |
+
"grad_norm": 0.41796875,
|
| 667 |
+
"learning_rate": 0.00019123006268411205,
|
| 668 |
+
"loss": 0.8363,
|
| 669 |
+
"step": 455
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.2276029055690074,
|
| 673 |
+
"grad_norm": 0.375,
|
| 674 |
+
"learning_rate": 0.00019087982683538114,
|
| 675 |
+
"loss": 0.8331,
|
| 676 |
+
"step": 460
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.25181598062954,
|
| 680 |
+
"grad_norm": 0.365234375,
|
| 681 |
+
"learning_rate": 0.00019052306741874532,
|
| 682 |
+
"loss": 0.8318,
|
| 683 |
+
"step": 465
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.2760290556900724,
|
| 687 |
+
"grad_norm": 0.259765625,
|
| 688 |
+
"learning_rate": 0.00019015981004323536,
|
| 689 |
+
"loss": 0.84,
|
| 690 |
+
"step": 470
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.3002421307506054,
|
| 694 |
+
"grad_norm": 0.26953125,
|
| 695 |
+
"learning_rate": 0.00018979008078432073,
|
| 696 |
+
"loss": 0.8364,
|
| 697 |
+
"step": 475
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.324455205811138,
|
| 701 |
+
"grad_norm": 0.326171875,
|
| 702 |
+
"learning_rate": 0.00018941390618203806,
|
| 703 |
+
"loss": 0.8352,
|
| 704 |
+
"step": 480
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.348668280871671,
|
| 708 |
+
"grad_norm": 0.296875,
|
| 709 |
+
"learning_rate": 0.00018903131323908578,
|
| 710 |
+
"loss": 0.8437,
|
| 711 |
+
"step": 485
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.3728813559322033,
|
| 715 |
+
"grad_norm": 0.322265625,
|
| 716 |
+
"learning_rate": 0.000188642329418886,
|
| 717 |
+
"loss": 0.8322,
|
| 718 |
+
"step": 490
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.3970944309927362,
|
| 722 |
+
"grad_norm": 0.34765625,
|
| 723 |
+
"learning_rate": 0.00018824698264361306,
|
| 724 |
+
"loss": 0.8349,
|
| 725 |
+
"step": 495
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.4213075060532687,
|
| 729 |
+
"grad_norm": 0.365234375,
|
| 730 |
+
"learning_rate": 0.0001878453012921891,
|
| 731 |
+
"loss": 0.8324,
|
| 732 |
+
"step": 500
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.4455205811138017,
|
| 736 |
+
"grad_norm": 0.375,
|
| 737 |
+
"learning_rate": 0.0001874373141982471,
|
| 738 |
+
"loss": 0.8378,
|
| 739 |
+
"step": 505
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 2.469733656174334,
|
| 743 |
+
"grad_norm": 0.28125,
|
| 744 |
+
"learning_rate": 0.00018702305064806108,
|
| 745 |
+
"loss": 0.8362,
|
| 746 |
+
"step": 510
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.4939467312348667,
|
| 750 |
+
"grad_norm": 0.248046875,
|
| 751 |
+
"learning_rate": 0.00018660254037844388,
|
| 752 |
+
"loss": 0.8287,
|
| 753 |
+
"step": 515
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 2.5181598062953996,
|
| 757 |
+
"grad_norm": 0.2431640625,
|
| 758 |
+
"learning_rate": 0.00018617581357461253,
|
| 759 |
+
"loss": 0.8367,
|
| 760 |
+
"step": 520
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.542372881355932,
|
| 764 |
+
"grad_norm": 0.318359375,
|
| 765 |
+
"learning_rate": 0.00018574290086802156,
|
| 766 |
+
"loss": 0.8317,
|
| 767 |
+
"step": 525
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 2.566585956416465,
|
| 771 |
+
"grad_norm": 0.2734375,
|
| 772 |
+
"learning_rate": 0.00018530383333416418,
|
| 773 |
+
"loss": 0.8361,
|
| 774 |
+
"step": 530
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.5907990314769975,
|
| 778 |
+
"grad_norm": 0.279296875,
|
| 779 |
+
"learning_rate": 0.00018485864249034146,
|
| 780 |
+
"loss": 0.8303,
|
| 781 |
+
"step": 535
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 2.61501210653753,
|
| 785 |
+
"grad_norm": 0.2890625,
|
| 786 |
+
"learning_rate": 0.00018440736029340022,
|
| 787 |
+
"loss": 0.8296,
|
| 788 |
+
"step": 540
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.639225181598063,
|
| 792 |
+
"grad_norm": 0.30078125,
|
| 793 |
+
"learning_rate": 0.00018395001913743892,
|
| 794 |
+
"loss": 0.8348,
|
| 795 |
+
"step": 545
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 2.663438256658596,
|
| 799 |
+
"grad_norm": 0.275390625,
|
| 800 |
+
"learning_rate": 0.00018348665185148228,
|
| 801 |
+
"loss": 0.8307,
|
| 802 |
+
"step": 550
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.6876513317191284,
|
| 806 |
+
"grad_norm": 0.34375,
|
| 807 |
+
"learning_rate": 0.0001830172916971249,
|
| 808 |
+
"loss": 0.8316,
|
| 809 |
+
"step": 555
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.711864406779661,
|
| 813 |
+
"grad_norm": 0.486328125,
|
| 814 |
+
"learning_rate": 0.00018254197236614354,
|
| 815 |
+
"loss": 0.8299,
|
| 816 |
+
"step": 560
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.736077481840194,
|
| 820 |
+
"grad_norm": 0.3125,
|
| 821 |
+
"learning_rate": 0.0001820607279780787,
|
| 822 |
+
"loss": 0.8278,
|
| 823 |
+
"step": 565
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 2.7602905569007263,
|
| 827 |
+
"grad_norm": 0.3359375,
|
| 828 |
+
"learning_rate": 0.00018157359307778545,
|
| 829 |
+
"loss": 0.8317,
|
| 830 |
+
"step": 570
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.7845036319612593,
|
| 834 |
+
"grad_norm": 0.2734375,
|
| 835 |
+
"learning_rate": 0.00018108060263295362,
|
| 836 |
+
"loss": 0.8246,
|
| 837 |
+
"step": 575
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 2.8087167070217918,
|
| 841 |
+
"grad_norm": 0.373046875,
|
| 842 |
+
"learning_rate": 0.00018058179203159788,
|
| 843 |
+
"loss": 0.8284,
|
| 844 |
+
"step": 580
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.8329297820823243,
|
| 848 |
+
"grad_norm": 0.30859375,
|
| 849 |
+
"learning_rate": 0.00018007719707951745,
|
| 850 |
+
"loss": 0.8273,
|
| 851 |
+
"step": 585
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 2.857142857142857,
|
| 855 |
+
"grad_norm": 0.279296875,
|
| 856 |
+
"learning_rate": 0.00017956685399772576,
|
| 857 |
+
"loss": 0.8227,
|
| 858 |
+
"step": 590
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.8813559322033897,
|
| 862 |
+
"grad_norm": 0.341796875,
|
| 863 |
+
"learning_rate": 0.00017905079941985064,
|
| 864 |
+
"loss": 0.8237,
|
| 865 |
+
"step": 595
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 2.9055690072639226,
|
| 869 |
+
"grad_norm": 0.259765625,
|
| 870 |
+
"learning_rate": 0.0001785290703895045,
|
| 871 |
+
"loss": 0.8243,
|
| 872 |
+
"step": 600
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.929782082324455,
|
| 876 |
+
"grad_norm": 0.25,
|
| 877 |
+
"learning_rate": 0.0001780017043576253,
|
| 878 |
+
"loss": 0.82,
|
| 879 |
+
"step": 605
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 2.9539951573849876,
|
| 883 |
+
"grad_norm": 0.271484375,
|
| 884 |
+
"learning_rate": 0.0001774687391797882,
|
| 885 |
+
"loss": 0.8223,
|
| 886 |
+
"step": 610
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.9782082324455206,
|
| 890 |
+
"grad_norm": 0.26953125,
|
| 891 |
+
"learning_rate": 0.00017693021311348837,
|
| 892 |
+
"loss": 0.8266,
|
| 893 |
+
"step": 615
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.9975786924939465,
|
| 897 |
+
"eval_loss": 2.6475303173065186,
|
| 898 |
+
"eval_runtime": 0.3372,
|
| 899 |
+
"eval_samples_per_second": 29.654,
|
| 900 |
+
"eval_steps_per_second": 2.965,
|
| 901 |
+
"step": 619
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.002421307506053,
|
| 905 |
+
"grad_norm": 0.30078125,
|
| 906 |
+
"learning_rate": 0.0001763861648153945,
|
| 907 |
+
"loss": 0.8135,
|
| 908 |
+
"step": 620
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 3.026634382566586,
|
| 912 |
+
"grad_norm": 0.244140625,
|
| 913 |
+
"learning_rate": 0.00017583663333857417,
|
| 914 |
+
"loss": 0.8029,
|
| 915 |
+
"step": 625
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 3.0508474576271185,
|
| 919 |
+
"grad_norm": 0.2294921875,
|
| 920 |
+
"learning_rate": 0.00017528165812969037,
|
| 921 |
+
"loss": 0.799,
|
| 922 |
+
"step": 630
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 3.0750605326876514,
|
| 926 |
+
"grad_norm": 0.31640625,
|
| 927 |
+
"learning_rate": 0.00017472127902617009,
|
| 928 |
+
"loss": 0.8072,
|
| 929 |
+
"step": 635
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.099273607748184,
|
| 933 |
+
"grad_norm": 0.3671875,
|
| 934 |
+
"learning_rate": 0.00017415553625334447,
|
| 935 |
+
"loss": 0.8003,
|
| 936 |
+
"step": 640
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.123486682808717,
|
| 940 |
+
"grad_norm": 0.291015625,
|
| 941 |
+
"learning_rate": 0.00017358447042156155,
|
| 942 |
+
"loss": 0.8047,
|
| 943 |
+
"step": 645
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.1476997578692494,
|
| 947 |
+
"grad_norm": 0.35546875,
|
| 948 |
+
"learning_rate": 0.00017300812252327104,
|
| 949 |
+
"loss": 0.8009,
|
| 950 |
+
"step": 650
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.171912832929782,
|
| 954 |
+
"grad_norm": 0.26953125,
|
| 955 |
+
"learning_rate": 0.00017242653393008183,
|
| 956 |
+
"loss": 0.8049,
|
| 957 |
+
"step": 655
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 3.196125907990315,
|
| 961 |
+
"grad_norm": 0.287109375,
|
| 962 |
+
"learning_rate": 0.00017183974638979222,
|
| 963 |
+
"loss": 0.8073,
|
| 964 |
+
"step": 660
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 3.2203389830508473,
|
| 968 |
+
"grad_norm": 0.357421875,
|
| 969 |
+
"learning_rate": 0.0001712478020233932,
|
| 970 |
+
"loss": 0.7974,
|
| 971 |
+
"step": 665
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.2445520581113803,
|
| 975 |
+
"grad_norm": 0.265625,
|
| 976 |
+
"learning_rate": 0.0001706507433220448,
|
| 977 |
+
"loss": 0.803,
|
| 978 |
+
"step": 670
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 3.2687651331719128,
|
| 982 |
+
"grad_norm": 0.353515625,
|
| 983 |
+
"learning_rate": 0.00017004861314402615,
|
| 984 |
+
"loss": 0.7959,
|
| 985 |
+
"step": 675
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.2929782082324457,
|
| 989 |
+
"grad_norm": 0.337890625,
|
| 990 |
+
"learning_rate": 0.0001694414547116588,
|
| 991 |
+
"loss": 0.8046,
|
| 992 |
+
"step": 680
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.317191283292978,
|
| 996 |
+
"grad_norm": 0.34375,
|
| 997 |
+
"learning_rate": 0.0001688293116082044,
|
| 998 |
+
"loss": 0.805,
|
| 999 |
+
"step": 685
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 3.341404358353511,
|
| 1003 |
+
"grad_norm": 0.35546875,
|
| 1004 |
+
"learning_rate": 0.00016821222777473583,
|
| 1005 |
+
"loss": 0.8041,
|
| 1006 |
+
"step": 690
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 3.3656174334140436,
|
| 1010 |
+
"grad_norm": 0.3125,
|
| 1011 |
+
"learning_rate": 0.00016759024750698335,
|
| 1012 |
+
"loss": 0.7956,
|
| 1013 |
+
"step": 695
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.389830508474576,
|
| 1017 |
+
"grad_norm": 0.26953125,
|
| 1018 |
+
"learning_rate": 0.00016696341545215475,
|
| 1019 |
+
"loss": 0.8002,
|
| 1020 |
+
"step": 700
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 3.414043583535109,
|
| 1024 |
+
"grad_norm": 0.27734375,
|
| 1025 |
+
"learning_rate": 0.00016633177660573047,
|
| 1026 |
+
"loss": 0.7987,
|
| 1027 |
+
"step": 705
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 3.4382566585956416,
|
| 1031 |
+
"grad_norm": 0.263671875,
|
| 1032 |
+
"learning_rate": 0.00016569537630823383,
|
| 1033 |
+
"loss": 0.8008,
|
| 1034 |
+
"step": 710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.4624697336561745,
|
| 1038 |
+
"grad_norm": 0.28515625,
|
| 1039 |
+
"learning_rate": 0.0001650542602419763,
|
| 1040 |
+
"loss": 0.7969,
|
| 1041 |
+
"step": 715
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 3.486682808716707,
|
| 1045 |
+
"grad_norm": 0.30078125,
|
| 1046 |
+
"learning_rate": 0.00016440847442777834,
|
| 1047 |
+
"loss": 0.7994,
|
| 1048 |
+
"step": 720
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 3.5108958837772395,
|
| 1052 |
+
"grad_norm": 0.318359375,
|
| 1053 |
+
"learning_rate": 0.00016375806522166586,
|
| 1054 |
+
"loss": 0.7991,
|
| 1055 |
+
"step": 725
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.5351089588377724,
|
| 1059 |
+
"grad_norm": 0.283203125,
|
| 1060 |
+
"learning_rate": 0.0001631030793115428,
|
| 1061 |
+
"loss": 0.8044,
|
| 1062 |
+
"step": 730
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 3.559322033898305,
|
| 1066 |
+
"grad_norm": 0.53515625,
|
| 1067 |
+
"learning_rate": 0.00016244356371383962,
|
| 1068 |
+
"loss": 0.796,
|
| 1069 |
+
"step": 735
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 3.583535108958838,
|
| 1073 |
+
"grad_norm": 0.279296875,
|
| 1074 |
+
"learning_rate": 0.00016177956577013847,
|
| 1075 |
+
"loss": 0.7946,
|
| 1076 |
+
"step": 740
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 3.6077481840193704,
|
| 1080 |
+
"grad_norm": 0.25390625,
|
| 1081 |
+
"learning_rate": 0.00016111113314377484,
|
| 1082 |
+
"loss": 0.8053,
|
| 1083 |
+
"step": 745
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 3.6319612590799033,
|
| 1087 |
+
"grad_norm": 0.251953125,
|
| 1088 |
+
"learning_rate": 0.00016043831381641616,
|
| 1089 |
+
"loss": 0.8015,
|
| 1090 |
+
"step": 750
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.656174334140436,
|
| 1094 |
+
"grad_norm": 0.283203125,
|
| 1095 |
+
"learning_rate": 0.00015976115608461758,
|
| 1096 |
+
"loss": 0.7998,
|
| 1097 |
+
"step": 755
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.6803874092009687,
|
| 1101 |
+
"grad_norm": 0.24609375,
|
| 1102 |
+
"learning_rate": 0.00015907970855635512,
|
| 1103 |
+
"loss": 0.8,
|
| 1104 |
+
"step": 760
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 3.7046004842615012,
|
| 1108 |
+
"grad_norm": 0.298828125,
|
| 1109 |
+
"learning_rate": 0.00015839402014753654,
|
| 1110 |
+
"loss": 0.7956,
|
| 1111 |
+
"step": 765
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 3.7288135593220337,
|
| 1115 |
+
"grad_norm": 0.296875,
|
| 1116 |
+
"learning_rate": 0.00015770414007848995,
|
| 1117 |
+
"loss": 0.7938,
|
| 1118 |
+
"step": 770
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 3.7530266343825667,
|
| 1122 |
+
"grad_norm": 0.25390625,
|
| 1123 |
+
"learning_rate": 0.00015701011787043077,
|
| 1124 |
+
"loss": 0.7987,
|
| 1125 |
+
"step": 775
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 3.777239709443099,
|
| 1129 |
+
"grad_norm": 0.29296875,
|
| 1130 |
+
"learning_rate": 0.00015631200334190675,
|
| 1131 |
+
"loss": 0.7965,
|
| 1132 |
+
"step": 780
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.801452784503632,
|
| 1136 |
+
"grad_norm": 0.302734375,
|
| 1137 |
+
"learning_rate": 0.00015560984660522223,
|
| 1138 |
+
"loss": 0.7904,
|
| 1139 |
+
"step": 785
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 3.8256658595641646,
|
| 1143 |
+
"grad_norm": 0.423828125,
|
| 1144 |
+
"learning_rate": 0.00015490369806284057,
|
| 1145 |
+
"loss": 0.7949,
|
| 1146 |
+
"step": 790
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 3.849878934624697,
|
| 1150 |
+
"grad_norm": 0.32421875,
|
| 1151 |
+
"learning_rate": 0.00015419360840376655,
|
| 1152 |
+
"loss": 0.7921,
|
| 1153 |
+
"step": 795
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 3.87409200968523,
|
| 1157 |
+
"grad_norm": 0.375,
|
| 1158 |
+
"learning_rate": 0.00015347962859990744,
|
| 1159 |
+
"loss": 0.7957,
|
| 1160 |
+
"step": 800
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 3.898305084745763,
|
| 1164 |
+
"grad_norm": 0.298828125,
|
| 1165 |
+
"learning_rate": 0.00015276180990241427,
|
| 1166 |
+
"loss": 0.7894,
|
| 1167 |
+
"step": 805
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 3.9225181598062955,
|
| 1171 |
+
"grad_norm": 0.36328125,
|
| 1172 |
+
"learning_rate": 0.00015204020383800292,
|
| 1173 |
+
"loss": 0.7923,
|
| 1174 |
+
"step": 810
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 3.946731234866828,
|
| 1178 |
+
"grad_norm": 0.384765625,
|
| 1179 |
+
"learning_rate": 0.00015131486220525543,
|
| 1180 |
+
"loss": 0.7907,
|
| 1181 |
+
"step": 815
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 3.970944309927361,
|
| 1185 |
+
"grad_norm": 0.283203125,
|
| 1186 |
+
"learning_rate": 0.00015058583707090164,
|
| 1187 |
+
"loss": 0.7921,
|
| 1188 |
+
"step": 820
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 3.9951573849878934,
|
| 1192 |
+
"grad_norm": 0.291015625,
|
| 1193 |
+
"learning_rate": 0.00014985318076608188,
|
| 1194 |
+
"loss": 0.7953,
|
| 1195 |
+
"step": 825
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 4.0,
|
| 1199 |
+
"eval_loss": 2.6602656841278076,
|
| 1200 |
+
"eval_runtime": 0.2371,
|
| 1201 |
+
"eval_samples_per_second": 42.178,
|
| 1202 |
+
"eval_steps_per_second": 4.218,
|
| 1203 |
+
"step": 826
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.019370460048426,
|
| 1207 |
+
"grad_norm": 0.29296875,
|
| 1208 |
+
"learning_rate": 0.00014911694588259038,
|
| 1209 |
+
"loss": 0.7705,
|
| 1210 |
+
"step": 830
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 4.043583535108959,
|
| 1214 |
+
"grad_norm": 0.291015625,
|
| 1215 |
+
"learning_rate": 0.00014837718526910018,
|
| 1216 |
+
"loss": 0.7729,
|
| 1217 |
+
"step": 835
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.067796610169491,
|
| 1221 |
+
"grad_norm": 0.279296875,
|
| 1222 |
+
"learning_rate": 0.0001476339520273696,
|
| 1223 |
+
"loss": 0.7692,
|
| 1224 |
+
"step": 840
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.092009685230024,
|
| 1228 |
+
"grad_norm": 0.294921875,
|
| 1229 |
+
"learning_rate": 0.00014688729950843035,
|
| 1230 |
+
"loss": 0.7757,
|
| 1231 |
+
"step": 845
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 4.116222760290557,
|
| 1235 |
+
"grad_norm": 0.32421875,
|
| 1236 |
+
"learning_rate": 0.0001461372813087578,
|
| 1237 |
+
"loss": 0.7793,
|
| 1238 |
+
"step": 850
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.14043583535109,
|
| 1242 |
+
"grad_norm": 0.27734375,
|
| 1243 |
+
"learning_rate": 0.00014538395126642394,
|
| 1244 |
+
"loss": 0.7784,
|
| 1245 |
+
"step": 855
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 4.164648910411622,
|
| 1249 |
+
"grad_norm": 0.3671875,
|
| 1250 |
+
"learning_rate": 0.0001446273634572326,
|
| 1251 |
+
"loss": 0.7735,
|
| 1252 |
+
"step": 860
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 4.188861985472155,
|
| 1256 |
+
"grad_norm": 0.296875,
|
| 1257 |
+
"learning_rate": 0.0001438675721908377,
|
| 1258 |
+
"loss": 0.7789,
|
| 1259 |
+
"step": 865
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 4.213075060532688,
|
| 1263 |
+
"grad_norm": 0.26171875,
|
| 1264 |
+
"learning_rate": 0.00014310463200684502,
|
| 1265 |
+
"loss": 0.7703,
|
| 1266 |
+
"step": 870
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 4.237288135593221,
|
| 1270 |
+
"grad_norm": 0.30859375,
|
| 1271 |
+
"learning_rate": 0.00014233859767089687,
|
| 1272 |
+
"loss": 0.775,
|
| 1273 |
+
"step": 875
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 4.261501210653753,
|
| 1277 |
+
"grad_norm": 0.30859375,
|
| 1278 |
+
"learning_rate": 0.00014156952417074115,
|
| 1279 |
+
"loss": 0.7792,
|
| 1280 |
+
"step": 880
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.285714285714286,
|
| 1284 |
+
"grad_norm": 0.314453125,
|
| 1285 |
+
"learning_rate": 0.00014079746671228418,
|
| 1286 |
+
"loss": 0.78,
|
| 1287 |
+
"step": 885
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 4.309927360774818,
|
| 1291 |
+
"grad_norm": 0.251953125,
|
| 1292 |
+
"learning_rate": 0.0001400224807156278,
|
| 1293 |
+
"loss": 0.774,
|
| 1294 |
+
"step": 890
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 4.3341404358353515,
|
| 1298 |
+
"grad_norm": 0.26171875,
|
| 1299 |
+
"learning_rate": 0.00013924462181109115,
|
| 1300 |
+
"loss": 0.7728,
|
| 1301 |
+
"step": 895
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 4.358353510895884,
|
| 1305 |
+
"grad_norm": 0.306640625,
|
| 1306 |
+
"learning_rate": 0.0001384639458352175,
|
| 1307 |
+
"loss": 0.7791,
|
| 1308 |
+
"step": 900
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 4.3825665859564165,
|
| 1312 |
+
"grad_norm": 0.279296875,
|
| 1313 |
+
"learning_rate": 0.0001376805088267662,
|
| 1314 |
+
"loss": 0.773,
|
| 1315 |
+
"step": 905
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 4.406779661016949,
|
| 1319 |
+
"grad_norm": 0.291015625,
|
| 1320 |
+
"learning_rate": 0.00013689436702268992,
|
| 1321 |
+
"loss": 0.7719,
|
| 1322 |
+
"step": 910
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.4309927360774815,
|
| 1326 |
+
"grad_norm": 0.2578125,
|
| 1327 |
+
"learning_rate": 0.00013610557685409796,
|
| 1328 |
+
"loss": 0.7744,
|
| 1329 |
+
"step": 915
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 4.455205811138015,
|
| 1333 |
+
"grad_norm": 0.3125,
|
| 1334 |
+
"learning_rate": 0.00013531419494220548,
|
| 1335 |
+
"loss": 0.7737,
|
| 1336 |
+
"step": 920
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 4.479418886198547,
|
| 1340 |
+
"grad_norm": 0.28125,
|
| 1341 |
+
"learning_rate": 0.00013452027809426903,
|
| 1342 |
+
"loss": 0.7787,
|
| 1343 |
+
"step": 925
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 4.50363196125908,
|
| 1347 |
+
"grad_norm": 0.302734375,
|
| 1348 |
+
"learning_rate": 0.0001337238832995089,
|
| 1349 |
+
"loss": 0.7687,
|
| 1350 |
+
"step": 930
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 4.527845036319612,
|
| 1354 |
+
"grad_norm": 0.259765625,
|
| 1355 |
+
"learning_rate": 0.00013292506772501819,
|
| 1356 |
+
"loss": 0.7721,
|
| 1357 |
+
"step": 935
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 4.552058111380145,
|
| 1361 |
+
"grad_norm": 0.30078125,
|
| 1362 |
+
"learning_rate": 0.00013212388871165927,
|
| 1363 |
+
"loss": 0.7715,
|
| 1364 |
+
"step": 940
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.576271186440678,
|
| 1368 |
+
"grad_norm": 0.28125,
|
| 1369 |
+
"learning_rate": 0.00013132040376994772,
|
| 1370 |
+
"loss": 0.7792,
|
| 1371 |
+
"step": 945
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 4.600484261501211,
|
| 1375 |
+
"grad_norm": 0.251953125,
|
| 1376 |
+
"learning_rate": 0.00013051467057592414,
|
| 1377 |
+
"loss": 0.7711,
|
| 1378 |
+
"step": 950
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 4.624697336561743,
|
| 1382 |
+
"grad_norm": 0.28515625,
|
| 1383 |
+
"learning_rate": 0.00012970674696701387,
|
| 1384 |
+
"loss": 0.7781,
|
| 1385 |
+
"step": 955
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 4.648910411622276,
|
| 1389 |
+
"grad_norm": 0.330078125,
|
| 1390 |
+
"learning_rate": 0.00012889669093787547,
|
| 1391 |
+
"loss": 0.7788,
|
| 1392 |
+
"step": 960
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 4.673123486682809,
|
| 1396 |
+
"grad_norm": 0.296875,
|
| 1397 |
+
"learning_rate": 0.0001280845606362376,
|
| 1398 |
+
"loss": 0.7768,
|
| 1399 |
+
"step": 965
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 4.697336561743342,
|
| 1403 |
+
"grad_norm": 0.31640625,
|
| 1404 |
+
"learning_rate": 0.0001272704143587251,
|
| 1405 |
+
"loss": 0.7722,
|
| 1406 |
+
"step": 970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.721549636803874,
|
| 1410 |
+
"grad_norm": 0.275390625,
|
| 1411 |
+
"learning_rate": 0.00012645431054667436,
|
| 1412 |
+
"loss": 0.771,
|
| 1413 |
+
"step": 975
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 4.745762711864407,
|
| 1417 |
+
"grad_norm": 0.259765625,
|
| 1418 |
+
"learning_rate": 0.00012563630778193805,
|
| 1419 |
+
"loss": 0.775,
|
| 1420 |
+
"step": 980
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 4.76997578692494,
|
| 1424 |
+
"grad_norm": 0.302734375,
|
| 1425 |
+
"learning_rate": 0.00012481646478268028,
|
| 1426 |
+
"loss": 0.7726,
|
| 1427 |
+
"step": 985
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 4.7941888619854724,
|
| 1431 |
+
"grad_norm": 0.36328125,
|
| 1432 |
+
"learning_rate": 0.00012399484039916147,
|
| 1433 |
+
"loss": 0.7687,
|
| 1434 |
+
"step": 990
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 4.818401937046005,
|
| 1438 |
+
"grad_norm": 0.279296875,
|
| 1439 |
+
"learning_rate": 0.00012317149360951407,
|
| 1440 |
+
"loss": 0.7716,
|
| 1441 |
+
"step": 995
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 4.842615012106537,
|
| 1445 |
+
"grad_norm": 0.275390625,
|
| 1446 |
+
"learning_rate": 0.0001223464835155087,
|
| 1447 |
+
"loss": 0.7717,
|
| 1448 |
+
"step": 1000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.86682808716707,
|
| 1452 |
+
"grad_norm": 0.291015625,
|
| 1453 |
+
"learning_rate": 0.00012151986933831213,
|
| 1454 |
+
"loss": 0.773,
|
| 1455 |
+
"step": 1005
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 4.891041162227603,
|
| 1459 |
+
"grad_norm": 0.2490234375,
|
| 1460 |
+
"learning_rate": 0.00012069171041423583,
|
| 1461 |
+
"loss": 0.7726,
|
| 1462 |
+
"step": 1010
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.915254237288136,
|
| 1466 |
+
"grad_norm": 0.25,
|
| 1467 |
+
"learning_rate": 0.00011986206619047693,
|
| 1468 |
+
"loss": 0.7721,
|
| 1469 |
+
"step": 1015
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 4.939467312348668,
|
| 1473 |
+
"grad_norm": 0.25390625,
|
| 1474 |
+
"learning_rate": 0.00011903099622085077,
|
| 1475 |
+
"loss": 0.7773,
|
| 1476 |
+
"step": 1020
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 4.963680387409201,
|
| 1480 |
+
"grad_norm": 0.296875,
|
| 1481 |
+
"learning_rate": 0.00011819856016151615,
|
| 1482 |
+
"loss": 0.776,
|
| 1483 |
+
"step": 1025
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 4.987893462469733,
|
| 1487 |
+
"grad_norm": 0.259765625,
|
| 1488 |
+
"learning_rate": 0.00011736481776669306,
|
| 1489 |
+
"loss": 0.7708,
|
| 1490 |
+
"step": 1030
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 4.9975786924939465,
|
| 1494 |
+
"eval_loss": 2.6720097064971924,
|
| 1495 |
+
"eval_runtime": 0.2452,
|
| 1496 |
+
"eval_samples_per_second": 40.776,
|
| 1497 |
+
"eval_steps_per_second": 4.078,
|
| 1498 |
+
"step": 1032
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 5.012106537530267,
|
| 1502 |
+
"grad_norm": 0.328125,
|
| 1503 |
+
"learning_rate": 0.00011652982888437316,
|
| 1504 |
+
"loss": 0.7584,
|
| 1505 |
+
"step": 1035
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 5.036319612590799,
|
| 1509 |
+
"grad_norm": 0.265625,
|
| 1510 |
+
"learning_rate": 0.00011569365345202414,
|
| 1511 |
+
"loss": 0.7465,
|
| 1512 |
+
"step": 1040
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 5.060532687651332,
|
| 1516 |
+
"grad_norm": 0.251953125,
|
| 1517 |
+
"learning_rate": 0.00011485635149228681,
|
| 1518 |
+
"loss": 0.7568,
|
| 1519 |
+
"step": 1045
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 5.084745762711864,
|
| 1523 |
+
"grad_norm": 0.318359375,
|
| 1524 |
+
"learning_rate": 0.00011401798310866693,
|
| 1525 |
+
"loss": 0.7503,
|
| 1526 |
+
"step": 1050
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 5.108958837772397,
|
| 1530 |
+
"grad_norm": 0.271484375,
|
| 1531 |
+
"learning_rate": 0.00011317860848122066,
|
| 1532 |
+
"loss": 0.7532,
|
| 1533 |
+
"step": 1055
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 5.13317191283293,
|
| 1537 |
+
"grad_norm": 0.271484375,
|
| 1538 |
+
"learning_rate": 0.00011233828786223464,
|
| 1539 |
+
"loss": 0.7531,
|
| 1540 |
+
"step": 1060
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 5.157384987893463,
|
| 1544 |
+
"grad_norm": 0.310546875,
|
| 1545 |
+
"learning_rate": 0.00011149708157190101,
|
| 1546 |
+
"loss": 0.7628,
|
| 1547 |
+
"step": 1065
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.181598062953995,
|
| 1551 |
+
"grad_norm": 0.2431640625,
|
| 1552 |
+
"learning_rate": 0.00011065504999398762,
|
| 1553 |
+
"loss": 0.7586,
|
| 1554 |
+
"step": 1070
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 5.2058111380145276,
|
| 1558 |
+
"grad_norm": 0.26953125,
|
| 1559 |
+
"learning_rate": 0.00010981225357150327,
|
| 1560 |
+
"loss": 0.7605,
|
| 1561 |
+
"step": 1075
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 5.230024213075061,
|
| 1565 |
+
"grad_norm": 0.267578125,
|
| 1566 |
+
"learning_rate": 0.00010896875280235907,
|
| 1567 |
+
"loss": 0.7627,
|
| 1568 |
+
"step": 1080
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 5.254237288135593,
|
| 1572 |
+
"grad_norm": 0.27734375,
|
| 1573 |
+
"learning_rate": 0.00010812460823502597,
|
| 1574 |
+
"loss": 0.7562,
|
| 1575 |
+
"step": 1085
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 5.278450363196126,
|
| 1579 |
+
"grad_norm": 0.275390625,
|
| 1580 |
+
"learning_rate": 0.00010727988046418809,
|
| 1581 |
+
"loss": 0.7571,
|
| 1582 |
+
"step": 1090
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 5.302663438256658,
|
| 1586 |
+
"grad_norm": 0.255859375,
|
| 1587 |
+
"learning_rate": 0.00010643463012639337,
|
| 1588 |
+
"loss": 0.7571,
|
| 1589 |
+
"step": 1095
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 5.326876513317191,
|
| 1593 |
+
"grad_norm": 0.251953125,
|
| 1594 |
+
"learning_rate": 0.00010558891789570082,
|
| 1595 |
+
"loss": 0.7583,
|
| 1596 |
+
"step": 1100
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 5.351089588377724,
|
| 1600 |
+
"grad_norm": 0.26953125,
|
| 1601 |
+
"learning_rate": 0.00010474280447932519,
|
| 1602 |
+
"loss": 0.7622,
|
| 1603 |
+
"step": 1105
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 5.375302663438257,
|
| 1607 |
+
"grad_norm": 0.3203125,
|
| 1608 |
+
"learning_rate": 0.00010389635061327925,
|
| 1609 |
+
"loss": 0.7554,
|
| 1610 |
+
"step": 1110
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 5.399515738498789,
|
| 1614 |
+
"grad_norm": 0.279296875,
|
| 1615 |
+
"learning_rate": 0.00010304961705801415,
|
| 1616 |
+
"loss": 0.7598,
|
| 1617 |
+
"step": 1115
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 5.423728813559322,
|
| 1621 |
+
"grad_norm": 0.265625,
|
| 1622 |
+
"learning_rate": 0.00010220266459405772,
|
| 1623 |
+
"loss": 0.7546,
|
| 1624 |
+
"step": 1120
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 5.447941888619855,
|
| 1628 |
+
"grad_norm": 0.283203125,
|
| 1629 |
+
"learning_rate": 0.00010135555401765153,
|
| 1630 |
+
"loss": 0.756,
|
| 1631 |
+
"step": 1125
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 5.472154963680388,
|
| 1635 |
+
"grad_norm": 0.296875,
|
| 1636 |
+
"learning_rate": 0.00010050834613638695,
|
| 1637 |
+
"loss": 0.7612,
|
| 1638 |
+
"step": 1130
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 5.49636803874092,
|
| 1642 |
+
"grad_norm": 0.275390625,
|
| 1643 |
+
"learning_rate": 9.966110176484004e-05,
|
| 1644 |
+
"loss": 0.7531,
|
| 1645 |
+
"step": 1135
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.520581113801453,
|
| 1649 |
+
"grad_norm": 0.2734375,
|
| 1650 |
+
"learning_rate": 9.881388172020624e-05,
|
| 1651 |
+
"loss": 0.7479,
|
| 1652 |
+
"step": 1140
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.544794188861985,
|
| 1656 |
+
"grad_norm": 0.287109375,
|
| 1657 |
+
"learning_rate": 9.796674681793471e-05,
|
| 1658 |
+
"loss": 0.76,
|
| 1659 |
+
"step": 1145
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 5.5690072639225185,
|
| 1663 |
+
"grad_norm": 0.255859375,
|
| 1664 |
+
"learning_rate": 9.711975786736296e-05,
|
| 1665 |
+
"loss": 0.7556,
|
| 1666 |
+
"step": 1150
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 5.593220338983051,
|
| 1670 |
+
"grad_norm": 0.2578125,
|
| 1671 |
+
"learning_rate": 9.627297566735167e-05,
|
| 1672 |
+
"loss": 0.7554,
|
| 1673 |
+
"step": 1155
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 5.6174334140435835,
|
| 1677 |
+
"grad_norm": 0.25390625,
|
| 1678 |
+
"learning_rate": 9.542646100192056e-05,
|
| 1679 |
+
"loss": 0.7617,
|
| 1680 |
+
"step": 1160
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 5.641646489104116,
|
| 1684 |
+
"grad_norm": 0.2734375,
|
| 1685 |
+
"learning_rate": 9.458027463588502e-05,
|
| 1686 |
+
"loss": 0.7589,
|
| 1687 |
+
"step": 1165
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 5.6658595641646485,
|
| 1691 |
+
"grad_norm": 0.26171875,
|
| 1692 |
+
"learning_rate": 9.373447731049436e-05,
|
| 1693 |
+
"loss": 0.7505,
|
| 1694 |
+
"step": 1170
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 5.690072639225182,
|
| 1698 |
+
"grad_norm": 0.30078125,
|
| 1699 |
+
"learning_rate": 9.288912973907166e-05,
|
| 1700 |
+
"loss": 0.7578,
|
| 1701 |
+
"step": 1175
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 5.714285714285714,
|
| 1705 |
+
"grad_norm": 0.341796875,
|
| 1706 |
+
"learning_rate": 9.20442926026555e-05,
|
| 1707 |
+
"loss": 0.7538,
|
| 1708 |
+
"step": 1180
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 5.738498789346247,
|
| 1712 |
+
"grad_norm": 0.28125,
|
| 1713 |
+
"learning_rate": 9.120002654564445e-05,
|
| 1714 |
+
"loss": 0.7563,
|
| 1715 |
+
"step": 1185
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 5.762711864406779,
|
| 1719 |
+
"grad_norm": 0.2734375,
|
| 1720 |
+
"learning_rate": 9.035639217144335e-05,
|
| 1721 |
+
"loss": 0.7517,
|
| 1722 |
+
"step": 1190
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 5.786924939467312,
|
| 1726 |
+
"grad_norm": 0.275390625,
|
| 1727 |
+
"learning_rate": 8.951345003811366e-05,
|
| 1728 |
+
"loss": 0.7513,
|
| 1729 |
+
"step": 1195
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 5.811138014527845,
|
| 1733 |
+
"grad_norm": 0.271484375,
|
| 1734 |
+
"learning_rate": 8.867126065402596e-05,
|
| 1735 |
+
"loss": 0.7614,
|
| 1736 |
+
"step": 1200
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 5.835351089588378,
|
| 1740 |
+
"grad_norm": 0.255859375,
|
| 1741 |
+
"learning_rate": 8.782988447351685e-05,
|
| 1742 |
+
"loss": 0.7584,
|
| 1743 |
+
"step": 1205
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 5.85956416464891,
|
| 1747 |
+
"grad_norm": 0.291015625,
|
| 1748 |
+
"learning_rate": 8.698938189254912e-05,
|
| 1749 |
+
"loss": 0.7596,
|
| 1750 |
+
"step": 1210
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 5.883777239709443,
|
| 1754 |
+
"grad_norm": 0.265625,
|
| 1755 |
+
"learning_rate": 8.614981324437667e-05,
|
| 1756 |
+
"loss": 0.7523,
|
| 1757 |
+
"step": 1215
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 5.907990314769976,
|
| 1761 |
+
"grad_norm": 0.283203125,
|
| 1762 |
+
"learning_rate": 8.53112387952134e-05,
|
| 1763 |
+
"loss": 0.7553,
|
| 1764 |
+
"step": 1220
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 5.932203389830509,
|
| 1768 |
+
"grad_norm": 0.302734375,
|
| 1769 |
+
"learning_rate": 8.447371873990736e-05,
|
| 1770 |
+
"loss": 0.7519,
|
| 1771 |
+
"step": 1225
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 5.956416464891041,
|
| 1775 |
+
"grad_norm": 0.310546875,
|
| 1776 |
+
"learning_rate": 8.363731319761964e-05,
|
| 1777 |
+
"loss": 0.7574,
|
| 1778 |
+
"step": 1230
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 5.980629539951574,
|
| 1782 |
+
"grad_norm": 0.265625,
|
| 1783 |
+
"learning_rate": 8.280208220750896e-05,
|
| 1784 |
+
"loss": 0.75,
|
| 1785 |
+
"step": 1235
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 6.0,
|
| 1789 |
+
"eval_loss": 2.6897997856140137,
|
| 1790 |
+
"eval_runtime": 0.2374,
|
| 1791 |
+
"eval_samples_per_second": 42.117,
|
| 1792 |
+
"eval_steps_per_second": 4.212,
|
| 1793 |
+
"step": 1239
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 6.004842615012106,
|
| 1797 |
+
"grad_norm": 0.27734375,
|
| 1798 |
+
"learning_rate": 8.196808572442203e-05,
|
| 1799 |
+
"loss": 0.7479,
|
| 1800 |
+
"step": 1240
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 6.0290556900726395,
|
| 1804 |
+
"grad_norm": 0.279296875,
|
| 1805 |
+
"learning_rate": 8.113538361458963e-05,
|
| 1806 |
+
"loss": 0.7448,
|
| 1807 |
+
"step": 1245
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 6.053268765133172,
|
| 1811 |
+
"grad_norm": 0.294921875,
|
| 1812 |
+
"learning_rate": 8.030403565132942e-05,
|
| 1813 |
+
"loss": 0.7442,
|
| 1814 |
+
"step": 1250
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.0774818401937045,
|
| 1818 |
+
"grad_norm": 0.255859375,
|
| 1819 |
+
"learning_rate": 7.94741015107552e-05,
|
| 1820 |
+
"loss": 0.7455,
|
| 1821 |
+
"step": 1255
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 6.101694915254237,
|
| 1825 |
+
"grad_norm": 0.263671875,
|
| 1826 |
+
"learning_rate": 7.864564076749335e-05,
|
| 1827 |
+
"loss": 0.7419,
|
| 1828 |
+
"step": 1260
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 6.12590799031477,
|
| 1832 |
+
"grad_norm": 0.26953125,
|
| 1833 |
+
"learning_rate": 7.781871289040618e-05,
|
| 1834 |
+
"loss": 0.7415,
|
| 1835 |
+
"step": 1265
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 6.150121065375303,
|
| 1839 |
+
"grad_norm": 0.255859375,
|
| 1840 |
+
"learning_rate": 7.699337723832337e-05,
|
| 1841 |
+
"loss": 0.7472,
|
| 1842 |
+
"step": 1270
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 6.174334140435835,
|
| 1846 |
+
"grad_norm": 0.26953125,
|
| 1847 |
+
"learning_rate": 7.616969305578084e-05,
|
| 1848 |
+
"loss": 0.7426,
|
| 1849 |
+
"step": 1275
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 6.198547215496368,
|
| 1853 |
+
"grad_norm": 0.255859375,
|
| 1854 |
+
"learning_rate": 7.53477194687683e-05,
|
| 1855 |
+
"loss": 0.7405,
|
| 1856 |
+
"step": 1280
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.2227602905569,
|
| 1860 |
+
"grad_norm": 0.259765625,
|
| 1861 |
+
"learning_rate": 7.452751548048471e-05,
|
| 1862 |
+
"loss": 0.7412,
|
| 1863 |
+
"step": 1285
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 6.246973365617434,
|
| 1867 |
+
"grad_norm": 0.26953125,
|
| 1868 |
+
"learning_rate": 7.370913996710316e-05,
|
| 1869 |
+
"loss": 0.7352,
|
| 1870 |
+
"step": 1290
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 6.271186440677966,
|
| 1874 |
+
"grad_norm": 0.322265625,
|
| 1875 |
+
"learning_rate": 7.289265167354449e-05,
|
| 1876 |
+
"loss": 0.7438,
|
| 1877 |
+
"step": 1295
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 6.295399515738499,
|
| 1881 |
+
"grad_norm": 0.28515625,
|
| 1882 |
+
"learning_rate": 7.20781092092604e-05,
|
| 1883 |
+
"loss": 0.7374,
|
| 1884 |
+
"step": 1300
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 6.319612590799031,
|
| 1888 |
+
"grad_norm": 0.25,
|
| 1889 |
+
"learning_rate": 7.126557104402655e-05,
|
| 1890 |
+
"loss": 0.7315,
|
| 1891 |
+
"step": 1305
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 6.343825665859564,
|
| 1895 |
+
"grad_norm": 0.259765625,
|
| 1896 |
+
"learning_rate": 7.045509550374509e-05,
|
| 1897 |
+
"loss": 0.7378,
|
| 1898 |
+
"step": 1310
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.368038740920097,
|
| 1902 |
+
"grad_norm": 0.28515625,
|
| 1903 |
+
"learning_rate": 6.964674076625827e-05,
|
| 1904 |
+
"loss": 0.7408,
|
| 1905 |
+
"step": 1315
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 6.39225181598063,
|
| 1909 |
+
"grad_norm": 0.318359375,
|
| 1910 |
+
"learning_rate": 6.884056485717198e-05,
|
| 1911 |
+
"loss": 0.7394,
|
| 1912 |
+
"step": 1320
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 6.416464891041162,
|
| 1916 |
+
"grad_norm": 0.2578125,
|
| 1917 |
+
"learning_rate": 6.80366256456908e-05,
|
| 1918 |
+
"loss": 0.7362,
|
| 1919 |
+
"step": 1325
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 6.440677966101695,
|
| 1923 |
+
"grad_norm": 0.251953125,
|
| 1924 |
+
"learning_rate": 6.723498084046378e-05,
|
| 1925 |
+
"loss": 0.7436,
|
| 1926 |
+
"step": 1330
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 6.464891041162228,
|
| 1930 |
+
"grad_norm": 0.2734375,
|
| 1931 |
+
"learning_rate": 6.64356879854422e-05,
|
| 1932 |
+
"loss": 0.7382,
|
| 1933 |
+
"step": 1335
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 6.4891041162227605,
|
| 1937 |
+
"grad_norm": 0.32421875,
|
| 1938 |
+
"learning_rate": 6.563880445574873e-05,
|
| 1939 |
+
"loss": 0.7433,
|
| 1940 |
+
"step": 1340
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.513317191283293,
|
| 1944 |
+
"grad_norm": 0.310546875,
|
| 1945 |
+
"learning_rate": 6.484438745355897e-05,
|
| 1946 |
+
"loss": 0.7443,
|
| 1947 |
+
"step": 1345
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 6.5375302663438255,
|
| 1951 |
+
"grad_norm": 0.275390625,
|
| 1952 |
+
"learning_rate": 6.405249400399554e-05,
|
| 1953 |
+
"loss": 0.7381,
|
| 1954 |
+
"step": 1350
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 6.561743341404358,
|
| 1958 |
+
"grad_norm": 0.322265625,
|
| 1959 |
+
"learning_rate": 6.326318095103432e-05,
|
| 1960 |
+
"loss": 0.7446,
|
| 1961 |
+
"step": 1355
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 6.585956416464891,
|
| 1965 |
+
"grad_norm": 0.29296875,
|
| 1966 |
+
"learning_rate": 6.247650495342435e-05,
|
| 1967 |
+
"loss": 0.7485,
|
| 1968 |
+
"step": 1360
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 6.610169491525424,
|
| 1972 |
+
"grad_norm": 0.3125,
|
| 1973 |
+
"learning_rate": 6.169252248062057e-05,
|
| 1974 |
+
"loss": 0.7444,
|
| 1975 |
+
"step": 1365
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 6.634382566585956,
|
| 1979 |
+
"grad_norm": 0.259765625,
|
| 1980 |
+
"learning_rate": 6.091128980873047e-05,
|
| 1981 |
+
"loss": 0.7421,
|
| 1982 |
+
"step": 1370
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 6.658595641646489,
|
| 1986 |
+
"grad_norm": 0.2890625,
|
| 1987 |
+
"learning_rate": 6.013286301647423e-05,
|
| 1988 |
+
"loss": 0.7514,
|
| 1989 |
+
"step": 1375
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 6.682808716707022,
|
| 1993 |
+
"grad_norm": 0.275390625,
|
| 1994 |
+
"learning_rate": 5.9357297981159497e-05,
|
| 1995 |
+
"loss": 0.7429,
|
| 1996 |
+
"step": 1380
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 6.707021791767555,
|
| 2000 |
+
"grad_norm": 0.259765625,
|
| 2001 |
+
"learning_rate": 5.8584650374670135e-05,
|
| 2002 |
+
"loss": 0.744,
|
| 2003 |
+
"step": 1385
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 6.731234866828087,
|
| 2007 |
+
"grad_norm": 0.259765625,
|
| 2008 |
+
"learning_rate": 5.7814975659470294e-05,
|
| 2009 |
+
"loss": 0.7445,
|
| 2010 |
+
"step": 1390
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 6.75544794188862,
|
| 2014 |
+
"grad_norm": 0.267578125,
|
| 2015 |
+
"learning_rate": 5.704832908462282e-05,
|
| 2016 |
+
"loss": 0.7528,
|
| 2017 |
+
"step": 1395
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 6.779661016949152,
|
| 2021 |
+
"grad_norm": 0.265625,
|
| 2022 |
+
"learning_rate": 5.628476568182349e-05,
|
| 2023 |
+
"loss": 0.7422,
|
| 2024 |
+
"step": 1400
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 6.803874092009686,
|
| 2028 |
+
"grad_norm": 0.25390625,
|
| 2029 |
+
"learning_rate": 5.552434026145097e-05,
|
| 2030 |
+
"loss": 0.7377,
|
| 2031 |
+
"step": 1405
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 6.828087167070218,
|
| 2035 |
+
"grad_norm": 0.26953125,
|
| 2036 |
+
"learning_rate": 5.476710740863189e-05,
|
| 2037 |
+
"loss": 0.7412,
|
| 2038 |
+
"step": 1410
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 6.852300242130751,
|
| 2042 |
+
"grad_norm": 0.259765625,
|
| 2043 |
+
"learning_rate": 5.4013121479323026e-05,
|
| 2044 |
+
"loss": 0.7445,
|
| 2045 |
+
"step": 1415
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 6.876513317191283,
|
| 2049 |
+
"grad_norm": 0.259765625,
|
| 2050 |
+
"learning_rate": 5.326243659640928e-05,
|
| 2051 |
+
"loss": 0.7414,
|
| 2052 |
+
"step": 1420
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 6.900726392251816,
|
| 2056 |
+
"grad_norm": 0.279296875,
|
| 2057 |
+
"learning_rate": 5.251510664581869e-05,
|
| 2058 |
+
"loss": 0.7455,
|
| 2059 |
+
"step": 1425
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 6.924939467312349,
|
| 2063 |
+
"grad_norm": 0.26953125,
|
| 2064 |
+
"learning_rate": 5.177118527265438e-05,
|
| 2065 |
+
"loss": 0.7436,
|
| 2066 |
+
"step": 1430
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 6.9491525423728815,
|
| 2070 |
+
"grad_norm": 0.248046875,
|
| 2071 |
+
"learning_rate": 5.1030725877343744e-05,
|
| 2072 |
+
"loss": 0.7402,
|
| 2073 |
+
"step": 1435
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 6.973365617433414,
|
| 2077 |
+
"grad_norm": 0.255859375,
|
| 2078 |
+
"learning_rate": 5.0293781611805146e-05,
|
| 2079 |
+
"loss": 0.7432,
|
| 2080 |
+
"step": 1440
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 6.9975786924939465,
|
| 2084 |
+
"grad_norm": 0.248046875,
|
| 2085 |
+
"learning_rate": 4.956040537563291e-05,
|
| 2086 |
+
"loss": 0.7446,
|
| 2087 |
+
"step": 1445
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 6.9975786924939465,
|
| 2091 |
+
"eval_loss": 2.702648639678955,
|
| 2092 |
+
"eval_runtime": 0.2447,
|
| 2093 |
+
"eval_samples_per_second": 40.862,
|
| 2094 |
+
"eval_steps_per_second": 4.086,
|
| 2095 |
+
"step": 1445
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.021791767554479,
|
| 2099 |
+
"grad_norm": 0.279296875,
|
| 2100 |
+
"learning_rate": 4.8830649812299555e-05,
|
| 2101 |
+
"loss": 0.7359,
|
| 2102 |
+
"step": 1450
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.046004842615012,
|
| 2106 |
+
"grad_norm": 0.32421875,
|
| 2107 |
+
"learning_rate": 4.8104567305377336e-05,
|
| 2108 |
+
"loss": 0.7284,
|
| 2109 |
+
"step": 1455
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.070217917675545,
|
| 2113 |
+
"grad_norm": 0.28515625,
|
| 2114 |
+
"learning_rate": 4.738220997477785e-05,
|
| 2115 |
+
"loss": 0.7293,
|
| 2116 |
+
"step": 1460
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.094430992736077,
|
| 2120 |
+
"grad_norm": 0.267578125,
|
| 2121 |
+
"learning_rate": 4.6663629673010636e-05,
|
| 2122 |
+
"loss": 0.7348,
|
| 2123 |
+
"step": 1465
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.11864406779661,
|
| 2127 |
+
"grad_norm": 0.267578125,
|
| 2128 |
+
"learning_rate": 4.594887798146147e-05,
|
| 2129 |
+
"loss": 0.7407,
|
| 2130 |
+
"step": 1470
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.142857142857143,
|
| 2134 |
+
"grad_norm": 0.29296875,
|
| 2135 |
+
"learning_rate": 4.523800620668921e-05,
|
| 2136 |
+
"loss": 0.7301,
|
| 2137 |
+
"step": 1475
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.167070217917676,
|
| 2141 |
+
"grad_norm": 0.248046875,
|
| 2142 |
+
"learning_rate": 4.4531065376743316e-05,
|
| 2143 |
+
"loss": 0.7354,
|
| 2144 |
+
"step": 1480
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.191283292978208,
|
| 2148 |
+
"grad_norm": 0.251953125,
|
| 2149 |
+
"learning_rate": 4.382810623750073e-05,
|
| 2150 |
+
"loss": 0.7351,
|
| 2151 |
+
"step": 1485
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.215496368038741,
|
| 2155 |
+
"grad_norm": 0.271484375,
|
| 2156 |
+
"learning_rate": 4.312917924902328e-05,
|
| 2157 |
+
"loss": 0.7316,
|
| 2158 |
+
"step": 1490
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.239709443099273,
|
| 2162 |
+
"grad_norm": 0.267578125,
|
| 2163 |
+
"learning_rate": 4.243433458193541e-05,
|
| 2164 |
+
"loss": 0.7325,
|
| 2165 |
+
"step": 1495
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.263922518159807,
|
| 2169 |
+
"grad_norm": 0.26171875,
|
| 2170 |
+
"learning_rate": 4.174362211382315e-05,
|
| 2171 |
+
"loss": 0.7304,
|
| 2172 |
+
"step": 1500
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.288135593220339,
|
| 2176 |
+
"grad_norm": 0.255859375,
|
| 2177 |
+
"learning_rate": 4.1057091425653306e-05,
|
| 2178 |
+
"loss": 0.7348,
|
| 2179 |
+
"step": 1505
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.312348668280872,
|
| 2183 |
+
"grad_norm": 0.2578125,
|
| 2184 |
+
"learning_rate": 4.037479179821485e-05,
|
| 2185 |
+
"loss": 0.7358,
|
| 2186 |
+
"step": 1510
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 7.336561743341404,
|
| 2190 |
+
"grad_norm": 0.255859375,
|
| 2191 |
+
"learning_rate": 3.969677220858123e-05,
|
| 2192 |
+
"loss": 0.7333,
|
| 2193 |
+
"step": 1515
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.3607748184019375,
|
| 2197 |
+
"grad_norm": 0.263671875,
|
| 2198 |
+
"learning_rate": 3.902308132659457e-05,
|
| 2199 |
+
"loss": 0.7344,
|
| 2200 |
+
"step": 1520
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 7.38498789346247,
|
| 2204 |
+
"grad_norm": 0.255859375,
|
| 2205 |
+
"learning_rate": 3.8353767511372406e-05,
|
| 2206 |
+
"loss": 0.7329,
|
| 2207 |
+
"step": 1525
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.4092009685230025,
|
| 2211 |
+
"grad_norm": 0.259765625,
|
| 2212 |
+
"learning_rate": 3.768887880783589e-05,
|
| 2213 |
+
"loss": 0.7388,
|
| 2214 |
+
"step": 1530
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 7.433414043583535,
|
| 2218 |
+
"grad_norm": 0.25,
|
| 2219 |
+
"learning_rate": 3.702846294326137e-05,
|
| 2220 |
+
"loss": 0.7419,
|
| 2221 |
+
"step": 1535
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 7.4576271186440675,
|
| 2225 |
+
"grad_norm": 0.26171875,
|
| 2226 |
+
"learning_rate": 3.637256732385423e-05,
|
| 2227 |
+
"loss": 0.7328,
|
| 2228 |
+
"step": 1540
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.481840193704601,
|
| 2232 |
+
"grad_norm": 0.26171875,
|
| 2233 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 2234 |
+
"loss": 0.7352,
|
| 2235 |
+
"step": 1545
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.506053268765133,
|
| 2239 |
+
"grad_norm": 0.255859375,
|
| 2240 |
+
"learning_rate": 3.507452481961495e-05,
|
| 2241 |
+
"loss": 0.733,
|
| 2242 |
+
"step": 1550
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 7.530266343825666,
|
| 2246 |
+
"grad_norm": 0.26171875,
|
| 2247 |
+
"learning_rate": 3.443247111132942e-05,
|
| 2248 |
+
"loss": 0.733,
|
| 2249 |
+
"step": 1555
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.554479418886198,
|
| 2253 |
+
"grad_norm": 0.2470703125,
|
| 2254 |
+
"learning_rate": 3.379512399461603e-05,
|
| 2255 |
+
"loss": 0.7331,
|
| 2256 |
+
"step": 1560
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 7.578692493946731,
|
| 2260 |
+
"grad_norm": 0.2578125,
|
| 2261 |
+
"learning_rate": 3.316252921975116e-05,
|
| 2262 |
+
"loss": 0.7309,
|
| 2263 |
+
"step": 1565
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 7.602905569007264,
|
| 2267 |
+
"grad_norm": 0.2431640625,
|
| 2268 |
+
"learning_rate": 3.2534732195876924e-05,
|
| 2269 |
+
"loss": 0.7362,
|
| 2270 |
+
"step": 1570
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 7.627118644067797,
|
| 2274 |
+
"grad_norm": 0.25390625,
|
| 2275 |
+
"learning_rate": 3.191177798774155e-05,
|
| 2276 |
+
"loss": 0.7334,
|
| 2277 |
+
"step": 1575
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 7.651331719128329,
|
| 2281 |
+
"grad_norm": 0.2578125,
|
| 2282 |
+
"learning_rate": 3.129371131246459e-05,
|
| 2283 |
+
"loss": 0.7329,
|
| 2284 |
+
"step": 1580
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 7.675544794188862,
|
| 2288 |
+
"grad_norm": 0.26171875,
|
| 2289 |
+
"learning_rate": 3.0680576536326856e-05,
|
| 2290 |
+
"loss": 0.7381,
|
| 2291 |
+
"step": 1585
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 7.699757869249394,
|
| 2295 |
+
"grad_norm": 0.25390625,
|
| 2296 |
+
"learning_rate": 3.007241767158604e-05,
|
| 2297 |
+
"loss": 0.7282,
|
| 2298 |
+
"step": 1590
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 7.723970944309928,
|
| 2302 |
+
"grad_norm": 0.26953125,
|
| 2303 |
+
"learning_rate": 2.9469278373316987e-05,
|
| 2304 |
+
"loss": 0.737,
|
| 2305 |
+
"step": 1595
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 7.74818401937046,
|
| 2309 |
+
"grad_norm": 0.25390625,
|
| 2310 |
+
"learning_rate": 2.887120193627838e-05,
|
| 2311 |
+
"loss": 0.734,
|
| 2312 |
+
"step": 1600
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 7.772397094430993,
|
| 2316 |
+
"grad_norm": 0.255859375,
|
| 2317 |
+
"learning_rate": 2.827823129180477e-05,
|
| 2318 |
+
"loss": 0.7312,
|
| 2319 |
+
"step": 1605
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 7.796610169491525,
|
| 2323 |
+
"grad_norm": 0.25390625,
|
| 2324 |
+
"learning_rate": 2.769040900472488e-05,
|
| 2325 |
+
"loss": 0.7346,
|
| 2326 |
+
"step": 1610
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 7.8208232445520585,
|
| 2330 |
+
"grad_norm": 0.25,
|
| 2331 |
+
"learning_rate": 2.7107777270306267e-05,
|
| 2332 |
+
"loss": 0.7406,
|
| 2333 |
+
"step": 1615
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 7.845036319612591,
|
| 2337 |
+
"grad_norm": 0.25390625,
|
| 2338 |
+
"learning_rate": 2.653037791122631e-05,
|
| 2339 |
+
"loss": 0.73,
|
| 2340 |
+
"step": 1620
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 7.8692493946731235,
|
| 2344 |
+
"grad_norm": 0.26171875,
|
| 2345 |
+
"learning_rate": 2.595825237457026e-05,
|
| 2346 |
+
"loss": 0.7352,
|
| 2347 |
+
"step": 1625
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 7.893462469733656,
|
| 2351 |
+
"grad_norm": 0.267578125,
|
| 2352 |
+
"learning_rate": 2.5391441728855925e-05,
|
| 2353 |
+
"loss": 0.7297,
|
| 2354 |
+
"step": 1630
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 7.917675544794189,
|
| 2358 |
+
"grad_norm": 0.244140625,
|
| 2359 |
+
"learning_rate": 2.482998666108577e-05,
|
| 2360 |
+
"loss": 0.7303,
|
| 2361 |
+
"step": 1635
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 7.941888619854722,
|
| 2365 |
+
"grad_norm": 0.26953125,
|
| 2366 |
+
"learning_rate": 2.427392747382623e-05,
|
| 2367 |
+
"loss": 0.7369,
|
| 2368 |
+
"step": 1640
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 7.966101694915254,
|
| 2372 |
+
"grad_norm": 0.263671875,
|
| 2373 |
+
"learning_rate": 2.372330408231479e-05,
|
| 2374 |
+
"loss": 0.7285,
|
| 2375 |
+
"step": 1645
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 7.990314769975787,
|
| 2379 |
+
"grad_norm": 0.25390625,
|
| 2380 |
+
"learning_rate": 2.3178156011594588e-05,
|
| 2381 |
+
"loss": 0.7301,
|
| 2382 |
+
"step": 1650
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 8.0,
|
| 2386 |
+
"eval_loss": 2.7081146240234375,
|
| 2387 |
+
"eval_runtime": 0.2373,
|
| 2388 |
+
"eval_samples_per_second": 42.146,
|
| 2389 |
+
"eval_steps_per_second": 4.215,
|
| 2390 |
+
"step": 1652
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 8.01452784503632,
|
| 2394 |
+
"grad_norm": 0.2578125,
|
| 2395 |
+
"learning_rate": 2.2638522393677563e-05,
|
| 2396 |
+
"loss": 0.7319,
|
| 2397 |
+
"step": 1655
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 8.038740920096853,
|
| 2401 |
+
"grad_norm": 0.255859375,
|
| 2402 |
+
"learning_rate": 2.2104441964735033e-05,
|
| 2403 |
+
"loss": 0.7268,
|
| 2404 |
+
"step": 1660
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 8.062953995157384,
|
| 2408 |
+
"grad_norm": 0.25390625,
|
| 2409 |
+
"learning_rate": 2.1575953062317477e-05,
|
| 2410 |
+
"loss": 0.7282,
|
| 2411 |
+
"step": 1665
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 8.087167070217918,
|
| 2415 |
+
"grad_norm": 0.2578125,
|
| 2416 |
+
"learning_rate": 2.1053093622602404e-05,
|
| 2417 |
+
"loss": 0.7299,
|
| 2418 |
+
"step": 1670
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 8.111380145278451,
|
| 2422 |
+
"grad_norm": 0.2470703125,
|
| 2423 |
+
"learning_rate": 2.0535901177671158e-05,
|
| 2424 |
+
"loss": 0.7253,
|
| 2425 |
+
"step": 1675
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 8.135593220338983,
|
| 2429 |
+
"grad_norm": 0.251953125,
|
| 2430 |
+
"learning_rate": 2.0024412852815032e-05,
|
| 2431 |
+
"loss": 0.7229,
|
| 2432 |
+
"step": 1680
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 8.159806295399516,
|
| 2436 |
+
"grad_norm": 0.2490234375,
|
| 2437 |
+
"learning_rate": 1.951866536386997e-05,
|
| 2438 |
+
"loss": 0.7292,
|
| 2439 |
+
"step": 1685
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 8.184019370460048,
|
| 2443 |
+
"grad_norm": 0.25,
|
| 2444 |
+
"learning_rate": 1.90186950145813e-05,
|
| 2445 |
+
"loss": 0.726,
|
| 2446 |
+
"step": 1690
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 8.208232445520581,
|
| 2450 |
+
"grad_norm": 0.25390625,
|
| 2451 |
+
"learning_rate": 1.8524537693997644e-05,
|
| 2452 |
+
"loss": 0.7343,
|
| 2453 |
+
"step": 1695
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 8.232445520581114,
|
| 2457 |
+
"grad_norm": 0.251953125,
|
| 2458 |
+
"learning_rate": 1.8036228873894746e-05,
|
| 2459 |
+
"loss": 0.7289,
|
| 2460 |
+
"step": 1700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 8.256658595641646,
|
| 2464 |
+
"grad_norm": 0.25,
|
| 2465 |
+
"learning_rate": 1.75538036062292e-05,
|
| 2466 |
+
"loss": 0.7272,
|
| 2467 |
+
"step": 1705
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 8.28087167070218,
|
| 2471 |
+
"grad_norm": 0.248046875,
|
| 2472 |
+
"learning_rate": 1.707729652062241e-05,
|
| 2473 |
+
"loss": 0.7366,
|
| 2474 |
+
"step": 1710
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 8.305084745762711,
|
| 2478 |
+
"grad_norm": 0.25,
|
| 2479 |
+
"learning_rate": 1.6606741821874595e-05,
|
| 2480 |
+
"loss": 0.7345,
|
| 2481 |
+
"step": 1715
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 8.329297820823244,
|
| 2485 |
+
"grad_norm": 0.24609375,
|
| 2486 |
+
"learning_rate": 1.6142173287509866e-05,
|
| 2487 |
+
"loss": 0.7367,
|
| 2488 |
+
"step": 1720
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 8.353510895883778,
|
| 2492 |
+
"grad_norm": 0.2451171875,
|
| 2493 |
+
"learning_rate": 1.5683624265351125e-05,
|
| 2494 |
+
"loss": 0.7297,
|
| 2495 |
+
"step": 1725
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 8.37772397094431,
|
| 2499 |
+
"grad_norm": 0.25,
|
| 2500 |
+
"learning_rate": 1.5231127671126677e-05,
|
| 2501 |
+
"loss": 0.7269,
|
| 2502 |
+
"step": 1730
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 8.401937046004843,
|
| 2506 |
+
"grad_norm": 0.2490234375,
|
| 2507 |
+
"learning_rate": 1.4784715986107267e-05,
|
| 2508 |
+
"loss": 0.7332,
|
| 2509 |
+
"step": 1735
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 8.426150121065376,
|
| 2513 |
+
"grad_norm": 0.25,
|
| 2514 |
+
"learning_rate": 1.4344421254774431e-05,
|
| 2515 |
+
"loss": 0.724,
|
| 2516 |
+
"step": 1740
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 8.450363196125908,
|
| 2520 |
+
"grad_norm": 0.25,
|
| 2521 |
+
"learning_rate": 1.3910275082520573e-05,
|
| 2522 |
+
"loss": 0.7241,
|
| 2523 |
+
"step": 1745
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 8.474576271186441,
|
| 2527 |
+
"grad_norm": 0.251953125,
|
| 2528 |
+
"learning_rate": 1.3482308633379837e-05,
|
| 2529 |
+
"loss": 0.7256,
|
| 2530 |
+
"step": 1750
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 8.498789346246973,
|
| 2534 |
+
"grad_norm": 0.251953125,
|
| 2535 |
+
"learning_rate": 1.3060552627791444e-05,
|
| 2536 |
+
"loss": 0.7281,
|
| 2537 |
+
"step": 1755
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 8.523002421307506,
|
| 2541 |
+
"grad_norm": 0.25390625,
|
| 2542 |
+
"learning_rate": 1.2645037340394284e-05,
|
| 2543 |
+
"loss": 0.7291,
|
| 2544 |
+
"step": 1760
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 8.54721549636804,
|
| 2548 |
+
"grad_norm": 0.25,
|
| 2549 |
+
"learning_rate": 1.2235792597853857e-05,
|
| 2550 |
+
"loss": 0.7321,
|
| 2551 |
+
"step": 1765
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 8.571428571428571,
|
| 2555 |
+
"grad_norm": 0.2578125,
|
| 2556 |
+
"learning_rate": 1.183284777672109e-05,
|
| 2557 |
+
"loss": 0.7317,
|
| 2558 |
+
"step": 1770
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 8.595641646489105,
|
| 2562 |
+
"grad_norm": 0.25390625,
|
| 2563 |
+
"learning_rate": 1.1436231801323882e-05,
|
| 2564 |
+
"loss": 0.7289,
|
| 2565 |
+
"step": 1775
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 8.619854721549636,
|
| 2569 |
+
"grad_norm": 0.2431640625,
|
| 2570 |
+
"learning_rate": 1.1045973141690513e-05,
|
| 2571 |
+
"loss": 0.73,
|
| 2572 |
+
"step": 1780
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 8.64406779661017,
|
| 2576 |
+
"grad_norm": 0.251953125,
|
| 2577 |
+
"learning_rate": 1.0662099811506332e-05,
|
| 2578 |
+
"loss": 0.7302,
|
| 2579 |
+
"step": 1785
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 8.668280871670703,
|
| 2583 |
+
"grad_norm": 0.255859375,
|
| 2584 |
+
"learning_rate": 1.02846393661026e-05,
|
| 2585 |
+
"loss": 0.7322,
|
| 2586 |
+
"step": 1790
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 8.692493946731235,
|
| 2590 |
+
"grad_norm": 0.2470703125,
|
| 2591 |
+
"learning_rate": 9.913618900478594e-06,
|
| 2592 |
+
"loss": 0.7335,
|
| 2593 |
+
"step": 1795
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 8.716707021791768,
|
| 2597 |
+
"grad_norm": 0.2470703125,
|
| 2598 |
+
"learning_rate": 9.549065047356832e-06,
|
| 2599 |
+
"loss": 0.7315,
|
| 2600 |
+
"step": 1800
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 8.7409200968523,
|
| 2604 |
+
"grad_norm": 0.25390625,
|
| 2605 |
+
"learning_rate": 9.191003975270985e-06,
|
| 2606 |
+
"loss": 0.7294,
|
| 2607 |
+
"step": 1805
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 8.765133171912833,
|
| 2611 |
+
"grad_norm": 0.2490234375,
|
| 2612 |
+
"learning_rate": 8.839461386687719e-06,
|
| 2613 |
+
"loss": 0.7313,
|
| 2614 |
+
"step": 1810
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 8.789346246973366,
|
| 2618 |
+
"grad_norm": 0.248046875,
|
| 2619 |
+
"learning_rate": 8.494462516161606e-06,
|
| 2620 |
+
"loss": 0.725,
|
| 2621 |
+
"step": 1815
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 8.813559322033898,
|
| 2625 |
+
"grad_norm": 0.25,
|
| 2626 |
+
"learning_rate": 8.156032128523694e-06,
|
| 2627 |
+
"loss": 0.7298,
|
| 2628 |
+
"step": 1820
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 8.837772397094431,
|
| 2632 |
+
"grad_norm": 0.2490234375,
|
| 2633 |
+
"learning_rate": 7.8241945171039e-06,
|
| 2634 |
+
"loss": 0.7303,
|
| 2635 |
+
"step": 1825
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 8.861985472154963,
|
| 2639 |
+
"grad_norm": 0.25,
|
| 2640 |
+
"learning_rate": 7.498973501987128e-06,
|
| 2641 |
+
"loss": 0.7286,
|
| 2642 |
+
"step": 1830
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 8.886198547215496,
|
| 2646 |
+
"grad_norm": 0.2490234375,
|
| 2647 |
+
"learning_rate": 7.180392428303395e-06,
|
| 2648 |
+
"loss": 0.7309,
|
| 2649 |
+
"step": 1835
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 8.91041162227603,
|
| 2653 |
+
"grad_norm": 0.25,
|
| 2654 |
+
"learning_rate": 6.868474164552185e-06,
|
| 2655 |
+
"loss": 0.7292,
|
| 2656 |
+
"step": 1840
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 8.934624697336561,
|
| 2660 |
+
"grad_norm": 0.2470703125,
|
| 2661 |
+
"learning_rate": 6.563241100960682e-06,
|
| 2662 |
+
"loss": 0.7356,
|
| 2663 |
+
"step": 1845
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 8.958837772397095,
|
| 2667 |
+
"grad_norm": 0.25,
|
| 2668 |
+
"learning_rate": 6.264715147876743e-06,
|
| 2669 |
+
"loss": 0.728,
|
| 2670 |
+
"step": 1850
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 8.983050847457626,
|
| 2674 |
+
"grad_norm": 0.248046875,
|
| 2675 |
+
"learning_rate": 5.9729177341960265e-06,
|
| 2676 |
+
"loss": 0.7268,
|
| 2677 |
+
"step": 1855
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 8.997578692493947,
|
| 2681 |
+
"eval_loss": 2.7127978801727295,
|
| 2682 |
+
"eval_runtime": 0.2436,
|
| 2683 |
+
"eval_samples_per_second": 41.055,
|
| 2684 |
+
"eval_steps_per_second": 4.105,
|
| 2685 |
+
"step": 1858
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 9.00726392251816,
|
| 2689 |
+
"grad_norm": 0.2578125,
|
| 2690 |
+
"learning_rate": 5.687869805823765e-06,
|
| 2691 |
+
"loss": 0.7189,
|
| 2692 |
+
"step": 1860
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 9.031476997578693,
|
| 2696 |
+
"grad_norm": 0.255859375,
|
| 2697 |
+
"learning_rate": 5.409591824171311e-06,
|
| 2698 |
+
"loss": 0.73,
|
| 2699 |
+
"step": 1865
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 9.055690072639225,
|
| 2703 |
+
"grad_norm": 0.26171875,
|
| 2704 |
+
"learning_rate": 5.13810376468723e-06,
|
| 2705 |
+
"loss": 0.7255,
|
| 2706 |
+
"step": 1870
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 9.079903147699758,
|
| 2710 |
+
"grad_norm": 0.24609375,
|
| 2711 |
+
"learning_rate": 4.8734251154235355e-06,
|
| 2712 |
+
"loss": 0.7256,
|
| 2713 |
+
"step": 1875
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 9.104116222760291,
|
| 2717 |
+
"grad_norm": 0.25,
|
| 2718 |
+
"learning_rate": 4.61557487563673e-06,
|
| 2719 |
+
"loss": 0.7276,
|
| 2720 |
+
"step": 1880
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 9.128329297820823,
|
| 2724 |
+
"grad_norm": 0.251953125,
|
| 2725 |
+
"learning_rate": 4.364571554424035e-06,
|
| 2726 |
+
"loss": 0.7269,
|
| 2727 |
+
"step": 1885
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 9.152542372881356,
|
| 2731 |
+
"grad_norm": 0.2490234375,
|
| 2732 |
+
"learning_rate": 4.1204331693946995e-06,
|
| 2733 |
+
"loss": 0.7322,
|
| 2734 |
+
"step": 1890
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 9.176755447941888,
|
| 2738 |
+
"grad_norm": 0.2578125,
|
| 2739 |
+
"learning_rate": 3.883177245376735e-06,
|
| 2740 |
+
"loss": 0.7291,
|
| 2741 |
+
"step": 1895
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 9.200968523002421,
|
| 2745 |
+
"grad_norm": 0.251953125,
|
| 2746 |
+
"learning_rate": 3.6528208131588526e-06,
|
| 2747 |
+
"loss": 0.7239,
|
| 2748 |
+
"step": 1900
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 9.225181598062955,
|
| 2752 |
+
"grad_norm": 0.2421875,
|
| 2753 |
+
"learning_rate": 3.4293804082680235e-06,
|
| 2754 |
+
"loss": 0.7319,
|
| 2755 |
+
"step": 1905
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 9.249394673123486,
|
| 2759 |
+
"grad_norm": 0.248046875,
|
| 2760 |
+
"learning_rate": 3.212872069782513e-06,
|
| 2761 |
+
"loss": 0.7287,
|
| 2762 |
+
"step": 1910
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 9.27360774818402,
|
| 2766 |
+
"grad_norm": 0.251953125,
|
| 2767 |
+
"learning_rate": 3.003311339180526e-06,
|
| 2768 |
+
"loss": 0.7316,
|
| 2769 |
+
"step": 1915
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 9.297820823244551,
|
| 2773 |
+
"grad_norm": 0.2490234375,
|
| 2774 |
+
"learning_rate": 2.800713259224619e-06,
|
| 2775 |
+
"loss": 0.732,
|
| 2776 |
+
"step": 1920
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 9.322033898305085,
|
| 2780 |
+
"grad_norm": 0.2470703125,
|
| 2781 |
+
"learning_rate": 2.6050923728818787e-06,
|
| 2782 |
+
"loss": 0.7283,
|
| 2783 |
+
"step": 1925
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 9.346246973365618,
|
| 2787 |
+
"grad_norm": 0.251953125,
|
| 2788 |
+
"learning_rate": 2.416462722280055e-06,
|
| 2789 |
+
"loss": 0.7281,
|
| 2790 |
+
"step": 1930
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 9.37046004842615,
|
| 2794 |
+
"grad_norm": 0.255859375,
|
| 2795 |
+
"learning_rate": 2.234837847699478e-06,
|
| 2796 |
+
"loss": 0.7308,
|
| 2797 |
+
"step": 1935
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 9.394673123486683,
|
| 2801 |
+
"grad_norm": 0.25390625,
|
| 2802 |
+
"learning_rate": 2.060230786601225e-06,
|
| 2803 |
+
"loss": 0.7277,
|
| 2804 |
+
"step": 1940
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 9.418886198547215,
|
| 2808 |
+
"grad_norm": 0.255859375,
|
| 2809 |
+
"learning_rate": 1.8926540726911469e-06,
|
| 2810 |
+
"loss": 0.7252,
|
| 2811 |
+
"step": 1945
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 9.443099273607748,
|
| 2815 |
+
"grad_norm": 0.255859375,
|
| 2816 |
+
"learning_rate": 1.7321197350202767e-06,
|
| 2817 |
+
"loss": 0.7328,
|
| 2818 |
+
"step": 1950
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 9.467312348668282,
|
| 2822 |
+
"grad_norm": 0.25390625,
|
| 2823 |
+
"learning_rate": 1.578639297121265e-06,
|
| 2824 |
+
"loss": 0.7235,
|
| 2825 |
+
"step": 1955
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 9.491525423728813,
|
| 2829 |
+
"grad_norm": 0.25,
|
| 2830 |
+
"learning_rate": 1.4322237761812408e-06,
|
| 2831 |
+
"loss": 0.73,
|
| 2832 |
+
"step": 1960
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 9.515738498789347,
|
| 2836 |
+
"grad_norm": 0.24609375,
|
| 2837 |
+
"learning_rate": 1.292883682250956e-06,
|
| 2838 |
+
"loss": 0.7251,
|
| 2839 |
+
"step": 1965
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 9.539951573849878,
|
| 2843 |
+
"grad_norm": 0.25390625,
|
| 2844 |
+
"learning_rate": 1.160629017490389e-06,
|
| 2845 |
+
"loss": 0.7299,
|
| 2846 |
+
"step": 1970
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 9.564164648910412,
|
| 2850 |
+
"grad_norm": 0.244140625,
|
| 2851 |
+
"learning_rate": 1.0354692754506845e-06,
|
| 2852 |
+
"loss": 0.7286,
|
| 2853 |
+
"step": 1975
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 9.588377723970945,
|
| 2857 |
+
"grad_norm": 0.2490234375,
|
| 2858 |
+
"learning_rate": 9.174134403927781e-07,
|
| 2859 |
+
"loss": 0.7318,
|
| 2860 |
+
"step": 1980
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 9.612590799031477,
|
| 2864 |
+
"grad_norm": 0.25390625,
|
| 2865 |
+
"learning_rate": 8.064699866424109e-07,
|
| 2866 |
+
"loss": 0.7284,
|
| 2867 |
+
"step": 1985
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 9.63680387409201,
|
| 2871 |
+
"grad_norm": 0.2490234375,
|
| 2872 |
+
"learning_rate": 7.026468779818385e-07,
|
| 2873 |
+
"loss": 0.7244,
|
| 2874 |
+
"step": 1990
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 9.661016949152543,
|
| 2878 |
+
"grad_norm": 0.25390625,
|
| 2879 |
+
"learning_rate": 6.059515670782334e-07,
|
| 2880 |
+
"loss": 0.737,
|
| 2881 |
+
"step": 1995
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 9.685230024213075,
|
| 2885 |
+
"grad_norm": 0.255859375,
|
| 2886 |
+
"learning_rate": 5.163909949486234e-07,
|
| 2887 |
+
"loss": 0.7309,
|
| 2888 |
+
"step": 2000
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 9.709443099273608,
|
| 2892 |
+
"grad_norm": 0.24609375,
|
| 2893 |
+
"learning_rate": 4.3397159046171254e-07,
|
| 2894 |
+
"loss": 0.7244,
|
| 2895 |
+
"step": 2005
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 9.73365617433414,
|
| 2899 |
+
"grad_norm": 0.24609375,
|
| 2900 |
+
"learning_rate": 3.5869926987636135e-07,
|
| 2901 |
+
"loss": 0.725,
|
| 2902 |
+
"step": 2010
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 9.757869249394673,
|
| 2906 |
+
"grad_norm": 0.251953125,
|
| 2907 |
+
"learning_rate": 2.9057943641693785e-07,
|
| 2908 |
+
"loss": 0.7296,
|
| 2909 |
+
"step": 2015
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 9.782082324455207,
|
| 2913 |
+
"grad_norm": 0.24609375,
|
| 2914 |
+
"learning_rate": 2.2961697988544972e-07,
|
| 2915 |
+
"loss": 0.7248,
|
| 2916 |
+
"step": 2020
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 9.806295399515738,
|
| 2920 |
+
"grad_norm": 0.251953125,
|
| 2921 |
+
"learning_rate": 1.7581627631053643e-07,
|
| 2922 |
+
"loss": 0.725,
|
| 2923 |
+
"step": 2025
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 9.830508474576272,
|
| 2927 |
+
"grad_norm": 0.25,
|
| 2928 |
+
"learning_rate": 1.2918118763335373e-07,
|
| 2929 |
+
"loss": 0.7326,
|
| 2930 |
+
"step": 2030
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 9.854721549636803,
|
| 2934 |
+
"grad_norm": 0.24609375,
|
| 2935 |
+
"learning_rate": 8.971506143032882e-08,
|
| 2936 |
+
"loss": 0.7304,
|
| 2937 |
+
"step": 2035
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 9.878934624697337,
|
| 2941 |
+
"grad_norm": 0.25390625,
|
| 2942 |
+
"learning_rate": 5.7420730672919086e-08,
|
| 2943 |
+
"loss": 0.7255,
|
| 2944 |
+
"step": 2040
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 9.90314769975787,
|
| 2948 |
+
"grad_norm": 0.255859375,
|
| 2949 |
+
"learning_rate": 3.230051352420826e-08,
|
| 2950 |
+
"loss": 0.7271,
|
| 2951 |
+
"step": 2045
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 9.927360774818402,
|
| 2955 |
+
"grad_norm": 0.25390625,
|
| 2956 |
+
"learning_rate": 1.435621317251723e-08,
|
| 2957 |
+
"loss": 0.7292,
|
| 2958 |
+
"step": 2050
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 9.951573849878935,
|
| 2962 |
+
"grad_norm": 0.24609375,
|
| 2963 |
+
"learning_rate": 3.5891177019520273e-09,
|
| 2964 |
+
"loss": 0.7296,
|
| 2965 |
+
"step": 2055
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 9.975786924939467,
|
| 2969 |
+
"grad_norm": 0.25390625,
|
| 2970 |
+
"learning_rate": 0.0,
|
| 2971 |
+
"loss": 0.7318,
|
| 2972 |
+
"step": 2060
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 9.975786924939467,
|
| 2976 |
+
"eval_loss": 2.712730884552002,
|
| 2977 |
+
"eval_runtime": 0.2369,
|
| 2978 |
+
"eval_samples_per_second": 42.216,
|
| 2979 |
+
"eval_steps_per_second": 4.222,
|
| 2980 |
+
"step": 2060
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 9.975786924939467,
|
| 2984 |
+
"step": 2060,
|
| 2985 |
+
"total_flos": 3.242523538363515e+18,
|
| 2986 |
+
"train_loss": 0.8949183881861492,
|
| 2987 |
+
"train_runtime": 11275.6837,
|
| 2988 |
+
"train_samples_per_second": 23.387,
|
| 2989 |
+
"train_steps_per_second": 0.183
|
| 2990 |
+
}
|
| 2991 |
+
],
|
| 2992 |
+
"logging_steps": 5,
|
| 2993 |
+
"max_steps": 2060,
|
| 2994 |
+
"num_input_tokens_seen": 0,
|
| 2995 |
+
"num_train_epochs": 10,
|
| 2996 |
+
"save_steps": 100,
|
| 2997 |
+
"stateful_callbacks": {
|
| 2998 |
+
"TrainerControl": {
|
| 2999 |
+
"args": {
|
| 3000 |
+
"should_epoch_stop": false,
|
| 3001 |
+
"should_evaluate": false,
|
| 3002 |
+
"should_log": false,
|
| 3003 |
+
"should_save": true,
|
| 3004 |
+
"should_training_stop": false
|
| 3005 |
+
},
|
| 3006 |
+
"attributes": {}
|
| 3007 |
+
}
|
| 3008 |
+
},
|
| 3009 |
+
"total_flos": 3.242523538363515e+18,
|
| 3010 |
+
"train_batch_size": 8,
|
| 3011 |
+
"trial_name": null,
|
| 3012 |
+
"trial_params": null
|
| 3013 |
+
}
|