Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: gemma
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: google/gemma-7b
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: gemma7b-summarize-11k
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# gemma7b-summarize-11k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.6219
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0002
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
+
- eval_batch_size: 2
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 2
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 16
|
| 50 |
+
- total_eval_batch_size: 4
|
| 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 |
+
| 2.8367 | 0.9907 | 53 | 2.7327 |
|
| 61 |
+
| 1.0189 | 2.0 | 107 | 2.3246 |
|
| 62 |
+
| 0.8788 | 2.9907 | 160 | 2.2786 |
|
| 63 |
+
| 0.8104 | 4.0 | 214 | 2.2754 |
|
| 64 |
+
| 0.7626 | 4.9907 | 267 | 2.3205 |
|
| 65 |
+
| 0.6665 | 6.0 | 321 | 2.3903 |
|
| 66 |
+
| 0.6015 | 6.9907 | 374 | 2.4630 |
|
| 67 |
+
| 0.5699 | 8.0 | 428 | 2.5837 |
|
| 68 |
+
| 0.5017 | 8.9907 | 481 | 2.6300 |
|
| 69 |
+
| 0.4969 | 9.9065 | 530 | 2.6219 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- PEFT 0.10.0
|
| 75 |
+
- Transformers 4.40.2
|
| 76 |
+
- Pytorch 2.3.0+cu121
|
| 77 |
+
- Datasets 2.19.1
|
| 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 25053936
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:456d9949002e1bcc86d1d09b7d11f883e840b0bb74b3b2d89f7c2ca1762e5dde
|
| 3 |
size 25053936
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.906542056074766,
|
| 3 |
+
"total_flos": 8.09000149933097e+17,
|
| 4 |
+
"train_loss": 2.438210007829486,
|
| 5 |
+
"train_runtime": 3271.6973,
|
| 6 |
+
"train_samples": 11013,
|
| 7 |
+
"train_samples_per_second": 2.613,
|
| 8 |
+
"train_steps_per_second": 0.162
|
| 9 |
+
}
|
runs/May14_19-17-37_deep-diver-main-sharp-rattlesnake-1-0-0/events.out.tfevents.1715728754.deep-diver-main-sharp-rattlesnake-1-0-0.385.0
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 sha256:9e089ae3d1d45ad57f0b209db2aba8afc8f87f1ab08fd445f8642bd57f3adaa6
|
| 3 |
+
size 30989
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.906542056074766,
|
| 3 |
+
"total_flos": 8.09000149933097e+17,
|
| 4 |
+
"train_loss": 2.438210007829486,
|
| 5 |
+
"train_runtime": 3271.6973,
|
| 6 |
+
"train_samples": 11013,
|
| 7 |
+
"train_samples_per_second": 2.613,
|
| 8 |
+
"train_steps_per_second": 0.162
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,859 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.906542056074766,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 530,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.018691588785046728,
|
| 13 |
+
"grad_norm": 456.0,
|
| 14 |
+
"learning_rate": 3.7735849056603773e-06,
|
| 15 |
+
"loss": 46.1216,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.09345794392523364,
|
| 20 |
+
"grad_norm": 438.0,
|
| 21 |
+
"learning_rate": 1.8867924528301888e-05,
|
| 22 |
+
"loss": 43.7935,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.18691588785046728,
|
| 27 |
+
"grad_norm": 302.0,
|
| 28 |
+
"learning_rate": 3.7735849056603776e-05,
|
| 29 |
+
"loss": 37.6584,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.2803738317757009,
|
| 34 |
+
"grad_norm": 41.0,
|
| 35 |
+
"learning_rate": 5.660377358490566e-05,
|
| 36 |
+
"loss": 21.0859,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.37383177570093457,
|
| 41 |
+
"grad_norm": 30.125,
|
| 42 |
+
"learning_rate": 7.547169811320755e-05,
|
| 43 |
+
"loss": 17.8237,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.4672897196261682,
|
| 48 |
+
"grad_norm": 14.5,
|
| 49 |
+
"learning_rate": 9.433962264150944e-05,
|
| 50 |
+
"loss": 15.4732,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.5607476635514018,
|
| 55 |
+
"grad_norm": 6.1875,
|
| 56 |
+
"learning_rate": 0.00011320754716981132,
|
| 57 |
+
"loss": 14.2199,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.6542056074766355,
|
| 62 |
+
"grad_norm": 6.8125,
|
| 63 |
+
"learning_rate": 0.0001320754716981132,
|
| 64 |
+
"loss": 13.4105,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.7476635514018691,
|
| 69 |
+
"grad_norm": 18.75,
|
| 70 |
+
"learning_rate": 0.0001509433962264151,
|
| 71 |
+
"loss": 12.0055,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.8411214953271028,
|
| 76 |
+
"grad_norm": 42.0,
|
| 77 |
+
"learning_rate": 0.000169811320754717,
|
| 78 |
+
"loss": 9.0376,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.9345794392523364,
|
| 83 |
+
"grad_norm": 5.125,
|
| 84 |
+
"learning_rate": 0.00018867924528301889,
|
| 85 |
+
"loss": 2.8367,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9906542056074766,
|
| 90 |
+
"eval_loss": 2.7326531410217285,
|
| 91 |
+
"eval_runtime": 1.0199,
|
| 92 |
+
"eval_samples_per_second": 4.902,
|
| 93 |
+
"eval_steps_per_second": 1.961,
|
| 94 |
+
"step": 53
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.02803738317757,
|
| 98 |
+
"grad_norm": 2.90625,
|
| 99 |
+
"learning_rate": 0.00019999132465602527,
|
| 100 |
+
"loss": 1.6892,
|
| 101 |
+
"step": 55
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.1214953271028036,
|
| 105 |
+
"grad_norm": 1.875,
|
| 106 |
+
"learning_rate": 0.0001998937443221316,
|
| 107 |
+
"loss": 1.4752,
|
| 108 |
+
"step": 60
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.2149532710280373,
|
| 112 |
+
"grad_norm": 1.3671875,
|
| 113 |
+
"learning_rate": 0.00019968784563700586,
|
| 114 |
+
"loss": 1.3666,
|
| 115 |
+
"step": 65
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.308411214953271,
|
| 119 |
+
"grad_norm": 1.171875,
|
| 120 |
+
"learning_rate": 0.00019937385186393888,
|
| 121 |
+
"loss": 1.251,
|
| 122 |
+
"step": 70
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.4018691588785046,
|
| 126 |
+
"grad_norm": 1.5078125,
|
| 127 |
+
"learning_rate": 0.0001989521034775823,
|
| 128 |
+
"loss": 1.2121,
|
| 129 |
+
"step": 75
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.4953271028037383,
|
| 133 |
+
"grad_norm": 1.7734375,
|
| 134 |
+
"learning_rate": 0.00019842305779475968,
|
| 135 |
+
"loss": 1.1493,
|
| 136 |
+
"step": 80
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.588785046728972,
|
| 140 |
+
"grad_norm": 6.40625,
|
| 141 |
+
"learning_rate": 0.0001977872884785815,
|
| 142 |
+
"loss": 1.149,
|
| 143 |
+
"step": 85
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.6822429906542056,
|
| 147 |
+
"grad_norm": 1.7109375,
|
| 148 |
+
"learning_rate": 0.00019704548491640192,
|
| 149 |
+
"loss": 1.0847,
|
| 150 |
+
"step": 90
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.7757009345794392,
|
| 154 |
+
"grad_norm": 2.75,
|
| 155 |
+
"learning_rate": 0.00019619845147229138,
|
| 156 |
+
"loss": 1.0808,
|
| 157 |
+
"step": 95
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.8691588785046729,
|
| 161 |
+
"grad_norm": 10.75,
|
| 162 |
+
"learning_rate": 0.00019524710661483592,
|
| 163 |
+
"loss": 1.011,
|
| 164 |
+
"step": 100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.9626168224299065,
|
| 168 |
+
"grad_norm": 5.25,
|
| 169 |
+
"learning_rate": 0.00019419248192120898,
|
| 170 |
+
"loss": 1.0189,
|
| 171 |
+
"step": 105
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.0,
|
| 175 |
+
"eval_loss": 2.324605703353882,
|
| 176 |
+
"eval_runtime": 1.0183,
|
| 177 |
+
"eval_samples_per_second": 4.91,
|
| 178 |
+
"eval_steps_per_second": 1.964,
|
| 179 |
+
"step": 107
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.05607476635514,
|
| 183 |
+
"grad_norm": 1.8046875,
|
| 184 |
+
"learning_rate": 0.00019303572095859546,
|
| 185 |
+
"loss": 0.964,
|
| 186 |
+
"step": 110
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.149532710280374,
|
| 190 |
+
"grad_norm": 1.9453125,
|
| 191 |
+
"learning_rate": 0.000191778078044181,
|
| 192 |
+
"loss": 0.9563,
|
| 193 |
+
"step": 115
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.2429906542056073,
|
| 197 |
+
"grad_norm": 2.5625,
|
| 198 |
+
"learning_rate": 0.00019042091688505105,
|
| 199 |
+
"loss": 0.9415,
|
| 200 |
+
"step": 120
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.336448598130841,
|
| 204 |
+
"grad_norm": 3.265625,
|
| 205 |
+
"learning_rate": 0.00018896570909947475,
|
| 206 |
+
"loss": 0.9123,
|
| 207 |
+
"step": 125
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 2.4299065420560746,
|
| 211 |
+
"grad_norm": 4.75,
|
| 212 |
+
"learning_rate": 0.00018741403262117657,
|
| 213 |
+
"loss": 0.9691,
|
| 214 |
+
"step": 130
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 2.5233644859813085,
|
| 218 |
+
"grad_norm": 0.93359375,
|
| 219 |
+
"learning_rate": 0.00018576756998832665,
|
| 220 |
+
"loss": 0.8991,
|
| 221 |
+
"step": 135
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.616822429906542,
|
| 225 |
+
"grad_norm": 2.71875,
|
| 226 |
+
"learning_rate": 0.00018402810651910442,
|
| 227 |
+
"loss": 0.8936,
|
| 228 |
+
"step": 140
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.710280373831776,
|
| 232 |
+
"grad_norm": 3.671875,
|
| 233 |
+
"learning_rate": 0.00018219752837581466,
|
| 234 |
+
"loss": 0.8844,
|
| 235 |
+
"step": 145
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.803738317757009,
|
| 239 |
+
"grad_norm": 1.859375,
|
| 240 |
+
"learning_rate": 0.00018027782051965407,
|
| 241 |
+
"loss": 0.9217,
|
| 242 |
+
"step": 150
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.897196261682243,
|
| 246 |
+
"grad_norm": 0.875,
|
| 247 |
+
"learning_rate": 0.00017827106455834733,
|
| 248 |
+
"loss": 0.9125,
|
| 249 |
+
"step": 155
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.9906542056074765,
|
| 253 |
+
"grad_norm": 0.74609375,
|
| 254 |
+
"learning_rate": 0.0001761794364889855,
|
| 255 |
+
"loss": 0.8788,
|
| 256 |
+
"step": 160
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.9906542056074765,
|
| 260 |
+
"eval_loss": 2.2786262035369873,
|
| 261 |
+
"eval_runtime": 1.0188,
|
| 262 |
+
"eval_samples_per_second": 4.908,
|
| 263 |
+
"eval_steps_per_second": 1.963,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 3.0841121495327104,
|
| 268 |
+
"grad_norm": 0.609375,
|
| 269 |
+
"learning_rate": 0.00017400520433851456,
|
| 270 |
+
"loss": 0.8796,
|
| 271 |
+
"step": 165
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 3.177570093457944,
|
| 275 |
+
"grad_norm": 1.0703125,
|
| 276 |
+
"learning_rate": 0.00017175072570443312,
|
| 277 |
+
"loss": 0.8224,
|
| 278 |
+
"step": 170
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 3.2710280373831777,
|
| 282 |
+
"grad_norm": 1.1953125,
|
| 283 |
+
"learning_rate": 0.00016941844519836508,
|
| 284 |
+
"loss": 0.8814,
|
| 285 |
+
"step": 175
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 3.364485981308411,
|
| 289 |
+
"grad_norm": 0.703125,
|
| 290 |
+
"learning_rate": 0.00016701089179528032,
|
| 291 |
+
"loss": 0.8648,
|
| 292 |
+
"step": 180
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 3.457943925233645,
|
| 296 |
+
"grad_norm": 1.1796875,
|
| 297 |
+
"learning_rate": 0.00016453067609123654,
|
| 298 |
+
"loss": 0.804,
|
| 299 |
+
"step": 185
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 3.5514018691588785,
|
| 303 |
+
"grad_norm": 0.70703125,
|
| 304 |
+
"learning_rate": 0.00016198048747261709,
|
| 305 |
+
"loss": 0.8115,
|
| 306 |
+
"step": 190
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 3.6448598130841123,
|
| 310 |
+
"grad_norm": 0.93359375,
|
| 311 |
+
"learning_rate": 0.0001593630911999333,
|
| 312 |
+
"loss": 0.7971,
|
| 313 |
+
"step": 195
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 3.7383177570093458,
|
| 317 |
+
"grad_norm": 1.3203125,
|
| 318 |
+
"learning_rate": 0.0001566813254093538,
|
| 319 |
+
"loss": 0.8242,
|
| 320 |
+
"step": 200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.831775700934579,
|
| 324 |
+
"grad_norm": 2.359375,
|
| 325 |
+
"learning_rate": 0.00015393809803521214,
|
| 326 |
+
"loss": 0.8288,
|
| 327 |
+
"step": 205
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 3.925233644859813,
|
| 331 |
+
"grad_norm": 2.0,
|
| 332 |
+
"learning_rate": 0.00015113638365682995,
|
| 333 |
+
"loss": 0.8104,
|
| 334 |
+
"step": 210
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 4.0,
|
| 338 |
+
"eval_loss": 2.2753615379333496,
|
| 339 |
+
"eval_runtime": 1.0175,
|
| 340 |
+
"eval_samples_per_second": 4.914,
|
| 341 |
+
"eval_steps_per_second": 1.966,
|
| 342 |
+
"step": 214
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 4.018691588785047,
|
| 346 |
+
"grad_norm": 1.078125,
|
| 347 |
+
"learning_rate": 0.00014827922027307451,
|
| 348 |
+
"loss": 0.82,
|
| 349 |
+
"step": 215
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 4.11214953271028,
|
| 353 |
+
"grad_norm": 0.8125,
|
| 354 |
+
"learning_rate": 0.00014536970600814787,
|
| 355 |
+
"loss": 0.7523,
|
| 356 |
+
"step": 220
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 4.205607476635514,
|
| 360 |
+
"grad_norm": 1.1171875,
|
| 361 |
+
"learning_rate": 0.0001424109957521806,
|
| 362 |
+
"loss": 0.772,
|
| 363 |
+
"step": 225
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 4.299065420560748,
|
| 367 |
+
"grad_norm": 1.515625,
|
| 368 |
+
"learning_rate": 0.0001394062977402717,
|
| 369 |
+
"loss": 0.7393,
|
| 370 |
+
"step": 230
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 4.392523364485982,
|
| 374 |
+
"grad_norm": 4.0625,
|
| 375 |
+
"learning_rate": 0.00013635887007368467,
|
| 376 |
+
"loss": 0.7492,
|
| 377 |
+
"step": 235
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 4.485981308411215,
|
| 381 |
+
"grad_norm": 1.3203125,
|
| 382 |
+
"learning_rate": 0.0001332720171869723,
|
| 383 |
+
"loss": 0.7598,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 4.579439252336448,
|
| 388 |
+
"grad_norm": 0.87109375,
|
| 389 |
+
"learning_rate": 0.0001301490862648603,
|
| 390 |
+
"loss": 0.7433,
|
| 391 |
+
"step": 245
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 4.672897196261682,
|
| 395 |
+
"grad_norm": 1.5,
|
| 396 |
+
"learning_rate": 0.0001269934636127754,
|
| 397 |
+
"loss": 0.7615,
|
| 398 |
+
"step": 250
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 4.766355140186916,
|
| 402 |
+
"grad_norm": 1.1875,
|
| 403 |
+
"learning_rate": 0.00012380857098495356,
|
| 404 |
+
"loss": 0.7309,
|
| 405 |
+
"step": 255
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 4.859813084112149,
|
| 409 |
+
"grad_norm": 0.921875,
|
| 410 |
+
"learning_rate": 0.00012059786187410984,
|
| 411 |
+
"loss": 0.7343,
|
| 412 |
+
"step": 260
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 4.953271028037383,
|
| 416 |
+
"grad_norm": 0.9296875,
|
| 417 |
+
"learning_rate": 0.00011736481776669306,
|
| 418 |
+
"loss": 0.7626,
|
| 419 |
+
"step": 265
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 4.990654205607477,
|
| 423 |
+
"eval_loss": 2.320525884628296,
|
| 424 |
+
"eval_runtime": 1.018,
|
| 425 |
+
"eval_samples_per_second": 4.912,
|
| 426 |
+
"eval_steps_per_second": 1.965,
|
| 427 |
+
"step": 267
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 5.046728971962617,
|
| 431 |
+
"grad_norm": 0.671875,
|
| 432 |
+
"learning_rate": 0.00011411294436778561,
|
| 433 |
+
"loss": 0.6804,
|
| 434 |
+
"step": 270
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 5.140186915887851,
|
| 438 |
+
"grad_norm": 0.7265625,
|
| 439 |
+
"learning_rate": 0.00011084576779974257,
|
| 440 |
+
"loss": 0.662,
|
| 441 |
+
"step": 275
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 5.233644859813084,
|
| 445 |
+
"grad_norm": 0.8203125,
|
| 446 |
+
"learning_rate": 0.00010756683077869132,
|
| 447 |
+
"loss": 0.6635,
|
| 448 |
+
"step": 280
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 5.327102803738318,
|
| 452 |
+
"grad_norm": 0.76953125,
|
| 453 |
+
"learning_rate": 0.00010427968877303808,
|
| 454 |
+
"loss": 0.6852,
|
| 455 |
+
"step": 285
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 5.420560747663552,
|
| 459 |
+
"grad_norm": 1.0546875,
|
| 460 |
+
"learning_rate": 0.00010098790614814658,
|
| 461 |
+
"loss": 0.6885,
|
| 462 |
+
"step": 290
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 5.5140186915887845,
|
| 466 |
+
"grad_norm": 1.046875,
|
| 467 |
+
"learning_rate": 9.769505230136962e-05,
|
| 468 |
+
"loss": 0.6665,
|
| 469 |
+
"step": 295
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 5.607476635514018,
|
| 473 |
+
"grad_norm": 0.9453125,
|
| 474 |
+
"learning_rate": 9.440469779162407e-05,
|
| 475 |
+
"loss": 0.6704,
|
| 476 |
+
"step": 300
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 5.700934579439252,
|
| 480 |
+
"grad_norm": 0.7890625,
|
| 481 |
+
"learning_rate": 9.112041046770653e-05,
|
| 482 |
+
"loss": 0.7145,
|
| 483 |
+
"step": 305
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 5.794392523364486,
|
| 487 |
+
"grad_norm": 0.8203125,
|
| 488 |
+
"learning_rate": 8.784575159954748e-05,
|
| 489 |
+
"loss": 0.6567,
|
| 490 |
+
"step": 310
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 5.88785046728972,
|
| 494 |
+
"grad_norm": 0.79296875,
|
| 495 |
+
"learning_rate": 8.458427201659927e-05,
|
| 496 |
+
"loss": 0.6752,
|
| 497 |
+
"step": 315
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 5.981308411214953,
|
| 501 |
+
"grad_norm": 1.046875,
|
| 502 |
+
"learning_rate": 8.13395082575451e-05,
|
| 503 |
+
"loss": 0.6665,
|
| 504 |
+
"step": 320
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 6.0,
|
| 508 |
+
"eval_loss": 2.3902792930603027,
|
| 509 |
+
"eval_runtime": 1.0171,
|
| 510 |
+
"eval_samples_per_second": 4.916,
|
| 511 |
+
"eval_steps_per_second": 1.966,
|
| 512 |
+
"step": 321
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 6.074766355140187,
|
| 516 |
+
"grad_norm": 0.859375,
|
| 517 |
+
"learning_rate": 7.81149787355039e-05,
|
| 518 |
+
"loss": 0.6076,
|
| 519 |
+
"step": 325
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 6.168224299065421,
|
| 523 |
+
"grad_norm": 1.0625,
|
| 524 |
+
"learning_rate": 7.491417992288927e-05,
|
| 525 |
+
"loss": 0.5788,
|
| 526 |
+
"step": 330
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 6.261682242990654,
|
| 530 |
+
"grad_norm": 0.8359375,
|
| 531 |
+
"learning_rate": 7.174058256006012e-05,
|
| 532 |
+
"loss": 0.6007,
|
| 533 |
+
"step": 335
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 6.355140186915888,
|
| 537 |
+
"grad_norm": 0.79296875,
|
| 538 |
+
"learning_rate": 6.859762789187259e-05,
|
| 539 |
+
"loss": 0.6096,
|
| 540 |
+
"step": 340
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 6.4485981308411215,
|
| 544 |
+
"grad_norm": 0.8671875,
|
| 545 |
+
"learning_rate": 6.548872393621578e-05,
|
| 546 |
+
"loss": 0.5978,
|
| 547 |
+
"step": 345
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 6.542056074766355,
|
| 551 |
+
"grad_norm": 0.79296875,
|
| 552 |
+
"learning_rate": 6.24172417885762e-05,
|
| 553 |
+
"loss": 0.6354,
|
| 554 |
+
"step": 350
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 6.635514018691588,
|
| 558 |
+
"grad_norm": 0.76171875,
|
| 559 |
+
"learning_rate": 5.938651196663865e-05,
|
| 560 |
+
"loss": 0.6094,
|
| 561 |
+
"step": 355
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 6.728971962616822,
|
| 565 |
+
"grad_norm": 0.859375,
|
| 566 |
+
"learning_rate": 5.639982079888726e-05,
|
| 567 |
+
"loss": 0.5985,
|
| 568 |
+
"step": 360
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 6.822429906542056,
|
| 572 |
+
"grad_norm": 0.765625,
|
| 573 |
+
"learning_rate": 5.3460406861121894e-05,
|
| 574 |
+
"loss": 0.5846,
|
| 575 |
+
"step": 365
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 6.91588785046729,
|
| 579 |
+
"grad_norm": 0.9140625,
|
| 580 |
+
"learning_rate": 5.0571457464755226e-05,
|
| 581 |
+
"loss": 0.6015,
|
| 582 |
+
"step": 370
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 6.990654205607477,
|
| 586 |
+
"eval_loss": 2.4630489349365234,
|
| 587 |
+
"eval_runtime": 1.0189,
|
| 588 |
+
"eval_samples_per_second": 4.907,
|
| 589 |
+
"eval_steps_per_second": 1.963,
|
| 590 |
+
"step": 374
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 7.009345794392523,
|
| 594 |
+
"grad_norm": 0.87109375,
|
| 595 |
+
"learning_rate": 4.7736105200697056e-05,
|
| 596 |
+
"loss": 0.6055,
|
| 597 |
+
"step": 375
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 7.102803738317757,
|
| 601 |
+
"grad_norm": 0.93359375,
|
| 602 |
+
"learning_rate": 4.495742454257418e-05,
|
| 603 |
+
"loss": 0.5464,
|
| 604 |
+
"step": 380
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 7.196261682242991,
|
| 608 |
+
"grad_norm": 0.93359375,
|
| 609 |
+
"learning_rate": 4.2238428512969064e-05,
|
| 610 |
+
"loss": 0.5556,
|
| 611 |
+
"step": 385
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 7.289719626168225,
|
| 615 |
+
"grad_norm": 0.83984375,
|
| 616 |
+
"learning_rate": 3.9582065416291926e-05,
|
| 617 |
+
"loss": 0.548,
|
| 618 |
+
"step": 390
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 7.383177570093458,
|
| 622 |
+
"grad_norm": 1.0390625,
|
| 623 |
+
"learning_rate": 3.69912156418289e-05,
|
| 624 |
+
"loss": 0.5074,
|
| 625 |
+
"step": 395
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 7.4766355140186915,
|
| 629 |
+
"grad_norm": 0.80859375,
|
| 630 |
+
"learning_rate": 3.4468688540433425e-05,
|
| 631 |
+
"loss": 0.5497,
|
| 632 |
+
"step": 400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 7.570093457943925,
|
| 636 |
+
"grad_norm": 0.8359375,
|
| 637 |
+
"learning_rate": 3.201721937824673e-05,
|
| 638 |
+
"loss": 0.5436,
|
| 639 |
+
"step": 405
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 7.663551401869158,
|
| 643 |
+
"grad_norm": 0.86328125,
|
| 644 |
+
"learning_rate": 2.9639466370751068e-05,
|
| 645 |
+
"loss": 0.5627,
|
| 646 |
+
"step": 410
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 7.757009345794392,
|
| 650 |
+
"grad_norm": 0.890625,
|
| 651 |
+
"learning_rate": 2.7338007800372024e-05,
|
| 652 |
+
"loss": 0.566,
|
| 653 |
+
"step": 415
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 7.850467289719626,
|
| 657 |
+
"grad_norm": 0.9296875,
|
| 658 |
+
"learning_rate": 2.5115339220754797e-05,
|
| 659 |
+
"loss": 0.557,
|
| 660 |
+
"step": 420
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 7.94392523364486,
|
| 664 |
+
"grad_norm": 0.84765625,
|
| 665 |
+
"learning_rate": 2.297387075074625e-05,
|
| 666 |
+
"loss": 0.5699,
|
| 667 |
+
"step": 425
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 8.0,
|
| 671 |
+
"eval_loss": 2.583679676055908,
|
| 672 |
+
"eval_runtime": 1.0173,
|
| 673 |
+
"eval_samples_per_second": 4.915,
|
| 674 |
+
"eval_steps_per_second": 1.966,
|
| 675 |
+
"step": 428
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 8.037383177570094,
|
| 679 |
+
"grad_norm": 0.796875,
|
| 680 |
+
"learning_rate": 2.09159244610172e-05,
|
| 681 |
+
"loss": 0.5156,
|
| 682 |
+
"step": 430
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 8.130841121495328,
|
| 686 |
+
"grad_norm": 0.828125,
|
| 687 |
+
"learning_rate": 1.89437318561583e-05,
|
| 688 |
+
"loss": 0.509,
|
| 689 |
+
"step": 435
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 8.22429906542056,
|
| 693 |
+
"grad_norm": 0.87109375,
|
| 694 |
+
"learning_rate": 1.7059431454979824e-05,
|
| 695 |
+
"loss": 0.5226,
|
| 696 |
+
"step": 440
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 8.317757009345794,
|
| 700 |
+
"grad_norm": 0.8046875,
|
| 701 |
+
"learning_rate": 1.52650664716397e-05,
|
| 702 |
+
"loss": 0.5088,
|
| 703 |
+
"step": 445
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 8.411214953271028,
|
| 707 |
+
"grad_norm": 0.82421875,
|
| 708 |
+
"learning_rate": 1.3562582600113294e-05,
|
| 709 |
+
"loss": 0.4948,
|
| 710 |
+
"step": 450
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 8.504672897196262,
|
| 714 |
+
"grad_norm": 0.8203125,
|
| 715 |
+
"learning_rate": 1.1953825904408034e-05,
|
| 716 |
+
"loss": 0.5357,
|
| 717 |
+
"step": 455
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 8.598130841121495,
|
| 721 |
+
"grad_norm": 0.8203125,
|
| 722 |
+
"learning_rate": 1.0440540816810395e-05,
|
| 723 |
+
"loss": 0.551,
|
| 724 |
+
"step": 460
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 8.69158878504673,
|
| 728 |
+
"grad_norm": 0.91796875,
|
| 729 |
+
"learning_rate": 9.024368246335735e-06,
|
| 730 |
+
"loss": 0.5092,
|
| 731 |
+
"step": 465
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 8.785046728971963,
|
| 735 |
+
"grad_norm": 0.83203125,
|
| 736 |
+
"learning_rate": 7.706843799431984e-06,
|
| 737 |
+
"loss": 0.5244,
|
| 738 |
+
"step": 470
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 8.878504672897197,
|
| 742 |
+
"grad_norm": 0.8671875,
|
| 743 |
+
"learning_rate": 6.489396114866941e-06,
|
| 744 |
+
"loss": 0.5235,
|
| 745 |
+
"step": 475
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 8.97196261682243,
|
| 749 |
+
"grad_norm": 0.8203125,
|
| 750 |
+
"learning_rate": 5.373345314604205e-06,
|
| 751 |
+
"loss": 0.5017,
|
| 752 |
+
"step": 480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 8.990654205607477,
|
| 756 |
+
"eval_loss": 2.6300225257873535,
|
| 757 |
+
"eval_runtime": 1.0187,
|
| 758 |
+
"eval_samples_per_second": 4.908,
|
| 759 |
+
"eval_steps_per_second": 1.963,
|
| 760 |
+
"step": 481
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 9.065420560747663,
|
| 764 |
+
"grad_norm": 0.81640625,
|
| 765 |
+
"learning_rate": 4.359901572347758e-06,
|
| 766 |
+
"loss": 0.5185,
|
| 767 |
+
"step": 485
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 9.158878504672897,
|
| 771 |
+
"grad_norm": 0.84765625,
|
| 772 |
+
"learning_rate": 3.450163801307582e-06,
|
| 773 |
+
"loss": 0.5366,
|
| 774 |
+
"step": 490
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 9.25233644859813,
|
| 778 |
+
"grad_norm": 0.796875,
|
| 779 |
+
"learning_rate": 2.6451184626087644e-06,
|
| 780 |
+
"loss": 0.5114,
|
| 781 |
+
"step": 495
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 9.345794392523365,
|
| 785 |
+
"grad_norm": 0.78125,
|
| 786 |
+
"learning_rate": 1.945638495636515e-06,
|
| 787 |
+
"loss": 0.5138,
|
| 788 |
+
"step": 500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 9.439252336448599,
|
| 792 |
+
"grad_norm": 0.8203125,
|
| 793 |
+
"learning_rate": 1.3524823714768375e-06,
|
| 794 |
+
"loss": 0.486,
|
| 795 |
+
"step": 505
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 9.532710280373832,
|
| 799 |
+
"grad_norm": 0.84375,
|
| 800 |
+
"learning_rate": 8.662932704792792e-07,
|
| 801 |
+
"loss": 0.5139,
|
| 802 |
+
"step": 510
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 9.626168224299064,
|
| 806 |
+
"grad_norm": 0.7890625,
|
| 807 |
+
"learning_rate": 4.875983848335874e-07,
|
| 808 |
+
"loss": 0.5092,
|
| 809 |
+
"step": 515
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 9.719626168224298,
|
| 813 |
+
"grad_norm": 0.7890625,
|
| 814 |
+
"learning_rate": 2.1680834691628626e-07,
|
| 815 |
+
"loss": 0.4957,
|
| 816 |
+
"step": 520
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 9.813084112149532,
|
| 820 |
+
"grad_norm": 0.8125,
|
| 821 |
+
"learning_rate": 5.421678402741659e-08,
|
| 822 |
+
"loss": 0.5253,
|
| 823 |
+
"step": 525
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 9.906542056074766,
|
| 827 |
+
"grad_norm": 0.82421875,
|
| 828 |
+
"learning_rate": 0.0,
|
| 829 |
+
"loss": 0.4969,
|
| 830 |
+
"step": 530
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 9.906542056074766,
|
| 834 |
+
"eval_loss": 2.6219420433044434,
|
| 835 |
+
"eval_runtime": 1.0178,
|
| 836 |
+
"eval_samples_per_second": 4.912,
|
| 837 |
+
"eval_steps_per_second": 1.965,
|
| 838 |
+
"step": 530
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 9.906542056074766,
|
| 842 |
+
"step": 530,
|
| 843 |
+
"total_flos": 8.09000149933097e+17,
|
| 844 |
+
"train_loss": 2.438210007829486,
|
| 845 |
+
"train_runtime": 3271.6973,
|
| 846 |
+
"train_samples_per_second": 2.613,
|
| 847 |
+
"train_steps_per_second": 0.162
|
| 848 |
+
}
|
| 849 |
+
],
|
| 850 |
+
"logging_steps": 5,
|
| 851 |
+
"max_steps": 530,
|
| 852 |
+
"num_input_tokens_seen": 0,
|
| 853 |
+
"num_train_epochs": 10,
|
| 854 |
+
"save_steps": 100,
|
| 855 |
+
"total_flos": 8.09000149933097e+17,
|
| 856 |
+
"train_batch_size": 4,
|
| 857 |
+
"trial_name": null,
|
| 858 |
+
"trial_params": null
|
| 859 |
+
}
|