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-2b
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: gemma2b-summarize-gpt4o-16k
|
| 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-summarize-gpt4o-16k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.5758
|
| 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: 8
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 3
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 48
|
| 50 |
+
- total_eval_batch_size: 24
|
| 51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
+
- lr_scheduler_type: cosine
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 54 |
+
- num_epochs: 10
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
+
| 1.8806 | 0.9863 | 36 | 2.6094 |
|
| 61 |
+
| 1.3239 | 2.0 | 73 | 2.5358 |
|
| 62 |
+
| 1.2327 | 2.9863 | 109 | 2.5192 |
|
| 63 |
+
| 1.1735 | 4.0 | 146 | 2.5203 |
|
| 64 |
+
| 1.1354 | 4.9863 | 182 | 2.5467 |
|
| 65 |
+
| 1.1015 | 6.0 | 219 | 2.5496 |
|
| 66 |
+
| 1.0858 | 6.9863 | 255 | 2.5680 |
|
| 67 |
+
| 1.0624 | 8.0 | 292 | 2.5723 |
|
| 68 |
+
| 1.0546 | 8.9863 | 328 | 2.5756 |
|
| 69 |
+
| 1.0623 | 9.8630 | 360 | 2.5758 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- PEFT 0.11.1
|
| 75 |
+
- Transformers 4.41.2
|
| 76 |
+
- Pytorch 2.3.0+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 19644912
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:283585d02f26a343c41d2c2dc9db52a2669fe9904094a966bd813779d536c6a6
|
| 3 |
size 19644912
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.863013698630137,
|
| 3 |
+
"total_flos": 2.1145440197646746e+17,
|
| 4 |
+
"train_loss": 1.288370986117257,
|
| 5 |
+
"train_runtime": 1930.7273,
|
| 6 |
+
"train_samples": 16152,
|
| 7 |
+
"train_samples_per_second": 9.074,
|
| 8 |
+
"train_steps_per_second": 0.186
|
| 9 |
+
}
|
runs/Jun05_14-12-12_user-HP-Z8-Fury-G5-Workstation-Desktop-PC/events.out.tfevents.1717564350.user-HP-Z8-Fury-G5-Workstation-Desktop-PC.25863.0
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:f8d1755fec369e9a066da439dc77107fe925342b73586691d4f238fccae494d8
|
| 3 |
+
size 23945
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.863013698630137,
|
| 3 |
+
"total_flos": 2.1145440197646746e+17,
|
| 4 |
+
"train_loss": 1.288370986117257,
|
| 5 |
+
"train_runtime": 1930.7273,
|
| 6 |
+
"train_samples": 16152,
|
| 7 |
+
"train_samples_per_second": 9.074,
|
| 8 |
+
"train_steps_per_second": 0.186
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,633 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.863013698630137,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 360,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0273972602739726,
|
| 13 |
+
"grad_norm": 1.8984375,
|
| 14 |
+
"learning_rate": 5.555555555555556e-06,
|
| 15 |
+
"loss": 3.0637,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.136986301369863,
|
| 20 |
+
"grad_norm": 2.265625,
|
| 21 |
+
"learning_rate": 2.777777777777778e-05,
|
| 22 |
+
"loss": 3.056,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.273972602739726,
|
| 27 |
+
"grad_norm": 2.015625,
|
| 28 |
+
"learning_rate": 5.555555555555556e-05,
|
| 29 |
+
"loss": 3.0171,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.410958904109589,
|
| 34 |
+
"grad_norm": 1.9609375,
|
| 35 |
+
"learning_rate": 8.333333333333334e-05,
|
| 36 |
+
"loss": 2.8054,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.547945205479452,
|
| 41 |
+
"grad_norm": 2.1875,
|
| 42 |
+
"learning_rate": 0.00011111111111111112,
|
| 43 |
+
"loss": 2.4802,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.684931506849315,
|
| 48 |
+
"grad_norm": 2.546875,
|
| 49 |
+
"learning_rate": 0.0001388888888888889,
|
| 50 |
+
"loss": 2.2946,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.821917808219178,
|
| 55 |
+
"grad_norm": 2.109375,
|
| 56 |
+
"learning_rate": 0.0001666666666666667,
|
| 57 |
+
"loss": 2.0875,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.958904109589041,
|
| 62 |
+
"grad_norm": 0.57421875,
|
| 63 |
+
"learning_rate": 0.00019444444444444446,
|
| 64 |
+
"loss": 1.8806,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.9863013698630136,
|
| 69 |
+
"eval_loss": 2.6093997955322266,
|
| 70 |
+
"eval_runtime": 0.5633,
|
| 71 |
+
"eval_samples_per_second": 17.753,
|
| 72 |
+
"eval_steps_per_second": 1.775,
|
| 73 |
+
"step": 36
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.095890410958904,
|
| 77 |
+
"grad_norm": 44.0,
|
| 78 |
+
"learning_rate": 0.00019992479525042303,
|
| 79 |
+
"loss": 1.6979,
|
| 80 |
+
"step": 40
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.2328767123287672,
|
| 84 |
+
"grad_norm": 0.62890625,
|
| 85 |
+
"learning_rate": 0.00019961946980917456,
|
| 86 |
+
"loss": 1.5925,
|
| 87 |
+
"step": 45
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.36986301369863,
|
| 91 |
+
"grad_norm": 0.4765625,
|
| 92 |
+
"learning_rate": 0.00019908004033648453,
|
| 93 |
+
"loss": 1.5152,
|
| 94 |
+
"step": 50
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.5068493150684932,
|
| 98 |
+
"grad_norm": 0.4765625,
|
| 99 |
+
"learning_rate": 0.00019830777448228603,
|
| 100 |
+
"loss": 1.4565,
|
| 101 |
+
"step": 55
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.643835616438356,
|
| 105 |
+
"grad_norm": 0.310546875,
|
| 106 |
+
"learning_rate": 0.00019730448705798239,
|
| 107 |
+
"loss": 1.3866,
|
| 108 |
+
"step": 60
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.7808219178082192,
|
| 112 |
+
"grad_norm": 0.31640625,
|
| 113 |
+
"learning_rate": 0.00019607253577167205,
|
| 114 |
+
"loss": 1.3522,
|
| 115 |
+
"step": 65
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.9178082191780823,
|
| 119 |
+
"grad_norm": 0.369140625,
|
| 120 |
+
"learning_rate": 0.00019461481568757506,
|
| 121 |
+
"loss": 1.3239,
|
| 122 |
+
"step": 70
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 2.0,
|
| 126 |
+
"eval_loss": 2.535839080810547,
|
| 127 |
+
"eval_runtime": 0.5617,
|
| 128 |
+
"eval_samples_per_second": 17.804,
|
| 129 |
+
"eval_steps_per_second": 1.78,
|
| 130 |
+
"step": 73
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0547945205479454,
|
| 134 |
+
"grad_norm": 0.26953125,
|
| 135 |
+
"learning_rate": 0.00019293475242268223,
|
| 136 |
+
"loss": 1.3035,
|
| 137 |
+
"step": 75
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.191780821917808,
|
| 141 |
+
"grad_norm": 0.2275390625,
|
| 142 |
+
"learning_rate": 0.0001910362940966147,
|
| 143 |
+
"loss": 1.2854,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.328767123287671,
|
| 148 |
+
"grad_norm": 0.306640625,
|
| 149 |
+
"learning_rate": 0.00018892390205361062,
|
| 150 |
+
"loss": 1.2525,
|
| 151 |
+
"step": 85
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.4657534246575343,
|
| 155 |
+
"grad_norm": 0.578125,
|
| 156 |
+
"learning_rate": 0.00018660254037844388,
|
| 157 |
+
"loss": 1.26,
|
| 158 |
+
"step": 90
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.602739726027397,
|
| 162 |
+
"grad_norm": 0.54296875,
|
| 163 |
+
"learning_rate": 0.00018407766423091034,
|
| 164 |
+
"loss": 1.2323,
|
| 165 |
+
"step": 95
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.73972602739726,
|
| 169 |
+
"grad_norm": 0.28515625,
|
| 170 |
+
"learning_rate": 0.00018135520702629675,
|
| 171 |
+
"loss": 1.2267,
|
| 172 |
+
"step": 100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8767123287671232,
|
| 176 |
+
"grad_norm": 0.267578125,
|
| 177 |
+
"learning_rate": 0.00017844156649195759,
|
| 178 |
+
"loss": 1.2327,
|
| 179 |
+
"step": 105
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.9863013698630136,
|
| 183 |
+
"eval_loss": 2.519155979156494,
|
| 184 |
+
"eval_runtime": 0.5663,
|
| 185 |
+
"eval_samples_per_second": 17.659,
|
| 186 |
+
"eval_steps_per_second": 1.766,
|
| 187 |
+
"step": 109
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 3.0136986301369864,
|
| 191 |
+
"grad_norm": 0.275390625,
|
| 192 |
+
"learning_rate": 0.00017534358963276607,
|
| 193 |
+
"loss": 1.2242,
|
| 194 |
+
"step": 110
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.1506849315068495,
|
| 198 |
+
"grad_norm": 0.30078125,
|
| 199 |
+
"learning_rate": 0.00017206855664077147,
|
| 200 |
+
"loss": 1.1938,
|
| 201 |
+
"step": 115
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.287671232876712,
|
| 205 |
+
"grad_norm": 0.51171875,
|
| 206 |
+
"learning_rate": 0.0001686241637868734,
|
| 207 |
+
"loss": 1.1879,
|
| 208 |
+
"step": 120
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 3.4246575342465753,
|
| 212 |
+
"grad_norm": 0.48828125,
|
| 213 |
+
"learning_rate": 0.00016501850533471836,
|
| 214 |
+
"loss": 1.1748,
|
| 215 |
+
"step": 125
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 3.5616438356164384,
|
| 219 |
+
"grad_norm": 0.29296875,
|
| 220 |
+
"learning_rate": 0.0001612600545193203,
|
| 221 |
+
"loss": 1.1704,
|
| 222 |
+
"step": 130
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 3.6986301369863015,
|
| 226 |
+
"grad_norm": 0.6015625,
|
| 227 |
+
"learning_rate": 0.0001573576436351046,
|
| 228 |
+
"loss": 1.1679,
|
| 229 |
+
"step": 135
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 3.8356164383561646,
|
| 233 |
+
"grad_norm": 0.392578125,
|
| 234 |
+
"learning_rate": 0.00015332044328016914,
|
| 235 |
+
"loss": 1.1663,
|
| 236 |
+
"step": 140
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 3.9726027397260273,
|
| 240 |
+
"grad_norm": 0.33203125,
|
| 241 |
+
"learning_rate": 0.00014915794080553707,
|
| 242 |
+
"loss": 1.1735,
|
| 243 |
+
"step": 145
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.0,
|
| 247 |
+
"eval_loss": 2.520301103591919,
|
| 248 |
+
"eval_runtime": 0.5527,
|
| 249 |
+
"eval_samples_per_second": 18.093,
|
| 250 |
+
"eval_steps_per_second": 1.809,
|
| 251 |
+
"step": 146
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 4.109589041095891,
|
| 255 |
+
"grad_norm": 0.302734375,
|
| 256 |
+
"learning_rate": 0.00014487991802004623,
|
| 257 |
+
"loss": 1.144,
|
| 258 |
+
"step": 150
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.2465753424657535,
|
| 262 |
+
"grad_norm": 0.310546875,
|
| 263 |
+
"learning_rate": 0.00014049642820326735,
|
| 264 |
+
"loss": 1.1273,
|
| 265 |
+
"step": 155
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 4.383561643835616,
|
| 269 |
+
"grad_norm": 0.51953125,
|
| 270 |
+
"learning_rate": 0.00013601777248047105,
|
| 271 |
+
"loss": 1.1344,
|
| 272 |
+
"step": 160
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 4.52054794520548,
|
| 276 |
+
"grad_norm": 0.6171875,
|
| 277 |
+
"learning_rate": 0.00013145447561516138,
|
| 278 |
+
"loss": 1.1364,
|
| 279 |
+
"step": 165
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.657534246575342,
|
| 283 |
+
"grad_norm": 0.50390625,
|
| 284 |
+
"learning_rate": 0.00012681726127606376,
|
| 285 |
+
"loss": 1.1344,
|
| 286 |
+
"step": 170
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 4.794520547945205,
|
| 290 |
+
"grad_norm": 0.453125,
|
| 291 |
+
"learning_rate": 0.00012211702683668878,
|
| 292 |
+
"loss": 1.1252,
|
| 293 |
+
"step": 175
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 4.931506849315069,
|
| 297 |
+
"grad_norm": 0.310546875,
|
| 298 |
+
"learning_rate": 0.00011736481776669306,
|
| 299 |
+
"loss": 1.1354,
|
| 300 |
+
"step": 180
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 4.986301369863014,
|
| 304 |
+
"eval_loss": 2.546712875366211,
|
| 305 |
+
"eval_runtime": 0.5597,
|
| 306 |
+
"eval_samples_per_second": 17.868,
|
| 307 |
+
"eval_steps_per_second": 1.787,
|
| 308 |
+
"step": 182
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 5.068493150684931,
|
| 312 |
+
"grad_norm": 0.419921875,
|
| 313 |
+
"learning_rate": 0.00011257180167521629,
|
| 314 |
+
"loss": 1.1069,
|
| 315 |
+
"step": 185
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 5.205479452054795,
|
| 319 |
+
"grad_norm": 0.515625,
|
| 320 |
+
"learning_rate": 0.0001077492420671931,
|
| 321 |
+
"loss": 1.0983,
|
| 322 |
+
"step": 190
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.342465753424658,
|
| 326 |
+
"grad_norm": 0.373046875,
|
| 327 |
+
"learning_rate": 0.00010290847187431113,
|
| 328 |
+
"loss": 1.0952,
|
| 329 |
+
"step": 195
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 5.47945205479452,
|
| 333 |
+
"grad_norm": 0.50390625,
|
| 334 |
+
"learning_rate": 9.806086682281758e-05,
|
| 335 |
+
"loss": 1.1105,
|
| 336 |
+
"step": 200
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 5.616438356164384,
|
| 340 |
+
"grad_norm": 0.470703125,
|
| 341 |
+
"learning_rate": 9.321781870075908e-05,
|
| 342 |
+
"loss": 1.1041,
|
| 343 |
+
"step": 205
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 5.7534246575342465,
|
| 347 |
+
"grad_norm": 0.4296875,
|
| 348 |
+
"learning_rate": 8.839070858747697e-05,
|
| 349 |
+
"loss": 1.0954,
|
| 350 |
+
"step": 210
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 5.890410958904109,
|
| 354 |
+
"grad_norm": 0.361328125,
|
| 355 |
+
"learning_rate": 8.35908801082676e-05,
|
| 356 |
+
"loss": 1.1015,
|
| 357 |
+
"step": 215
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 6.0,
|
| 361 |
+
"eval_loss": 2.5496411323547363,
|
| 362 |
+
"eval_runtime": 0.5525,
|
| 363 |
+
"eval_samples_per_second": 18.099,
|
| 364 |
+
"eval_steps_per_second": 1.81,
|
| 365 |
+
"step": 219
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 6.027397260273973,
|
| 369 |
+
"grad_norm": 0.35546875,
|
| 370 |
+
"learning_rate": 7.882961277705895e-05,
|
| 371 |
+
"loss": 1.0967,
|
| 372 |
+
"step": 220
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 6.164383561643835,
|
| 376 |
+
"grad_norm": 0.369140625,
|
| 377 |
+
"learning_rate": 7.411809548974792e-05,
|
| 378 |
+
"loss": 1.0792,
|
| 379 |
+
"step": 225
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 6.301369863013699,
|
| 383 |
+
"grad_norm": 0.4140625,
|
| 384 |
+
"learning_rate": 6.94674002304887e-05,
|
| 385 |
+
"loss": 1.0787,
|
| 386 |
+
"step": 230
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 6.438356164383562,
|
| 390 |
+
"grad_norm": 0.357421875,
|
| 391 |
+
"learning_rate": 6.488845605272113e-05,
|
| 392 |
+
"loss": 1.0699,
|
| 393 |
+
"step": 235
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 6.575342465753424,
|
| 397 |
+
"grad_norm": 0.384765625,
|
| 398 |
+
"learning_rate": 6.039202339608432e-05,
|
| 399 |
+
"loss": 1.0875,
|
| 400 |
+
"step": 240
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 6.712328767123288,
|
| 404 |
+
"grad_norm": 0.39453125,
|
| 405 |
+
"learning_rate": 5.5988668799569545e-05,
|
| 406 |
+
"loss": 1.0721,
|
| 407 |
+
"step": 245
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 6.8493150684931505,
|
| 411 |
+
"grad_norm": 0.33984375,
|
| 412 |
+
"learning_rate": 5.168874007033615e-05,
|
| 413 |
+
"loss": 1.0769,
|
| 414 |
+
"step": 250
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 6.986301369863014,
|
| 418 |
+
"grad_norm": 0.408203125,
|
| 419 |
+
"learning_rate": 4.7502341966544e-05,
|
| 420 |
+
"loss": 1.0858,
|
| 421 |
+
"step": 255
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 6.986301369863014,
|
| 425 |
+
"eval_loss": 2.568040609359741,
|
| 426 |
+
"eval_runtime": 0.5489,
|
| 427 |
+
"eval_samples_per_second": 18.22,
|
| 428 |
+
"eval_steps_per_second": 1.822,
|
| 429 |
+
"step": 255
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.123287671232877,
|
| 433 |
+
"grad_norm": 0.36328125,
|
| 434 |
+
"learning_rate": 4.343931245134616e-05,
|
| 435 |
+
"loss": 1.0577,
|
| 436 |
+
"step": 260
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.260273972602739,
|
| 440 |
+
"grad_norm": 0.515625,
|
| 441 |
+
"learning_rate": 3.950919957384582e-05,
|
| 442 |
+
"loss": 1.0614,
|
| 443 |
+
"step": 265
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.397260273972603,
|
| 447 |
+
"grad_norm": 0.388671875,
|
| 448 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 449 |
+
"loss": 1.0698,
|
| 450 |
+
"step": 270
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.534246575342466,
|
| 454 |
+
"grad_norm": 0.42578125,
|
| 455 |
+
"learning_rate": 3.2084332465620694e-05,
|
| 456 |
+
"loss": 1.0695,
|
| 457 |
+
"step": 275
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.671232876712329,
|
| 461 |
+
"grad_norm": 0.408203125,
|
| 462 |
+
"learning_rate": 2.8607026544210114e-05,
|
| 463 |
+
"loss": 1.0702,
|
| 464 |
+
"step": 280
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.808219178082192,
|
| 468 |
+
"grad_norm": 0.392578125,
|
| 469 |
+
"learning_rate": 2.529749287590042e-05,
|
| 470 |
+
"loss": 1.0632,
|
| 471 |
+
"step": 285
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.945205479452055,
|
| 475 |
+
"grad_norm": 0.40625,
|
| 476 |
+
"learning_rate": 2.2163508807583998e-05,
|
| 477 |
+
"loss": 1.0624,
|
| 478 |
+
"step": 290
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.0,
|
| 482 |
+
"eval_loss": 2.572319746017456,
|
| 483 |
+
"eval_runtime": 0.5428,
|
| 484 |
+
"eval_samples_per_second": 18.424,
|
| 485 |
+
"eval_steps_per_second": 1.842,
|
| 486 |
+
"step": 292
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 8.082191780821917,
|
| 490 |
+
"grad_norm": 0.359375,
|
| 491 |
+
"learning_rate": 1.921243914762889e-05,
|
| 492 |
+
"loss": 1.0583,
|
| 493 |
+
"step": 295
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.219178082191782,
|
| 497 |
+
"grad_norm": 0.33203125,
|
| 498 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 499 |
+
"loss": 1.0536,
|
| 500 |
+
"step": 300
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 8.356164383561644,
|
| 504 |
+
"grad_norm": 0.359375,
|
| 505 |
+
"learning_rate": 1.388633676074862e-05,
|
| 506 |
+
"loss": 1.0613,
|
| 507 |
+
"step": 305
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.493150684931507,
|
| 511 |
+
"grad_norm": 0.353515625,
|
| 512 |
+
"learning_rate": 1.1523820282334219e-05,
|
| 513 |
+
"loss": 1.0589,
|
| 514 |
+
"step": 310
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 8.63013698630137,
|
| 518 |
+
"grad_norm": 0.337890625,
|
| 519 |
+
"learning_rate": 9.369221296335006e-06,
|
| 520 |
+
"loss": 1.0664,
|
| 521 |
+
"step": 315
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 8.767123287671232,
|
| 525 |
+
"grad_norm": 0.337890625,
|
| 526 |
+
"learning_rate": 7.427603073110967e-06,
|
| 527 |
+
"loss": 1.0656,
|
| 528 |
+
"step": 320
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 8.904109589041095,
|
| 532 |
+
"grad_norm": 0.330078125,
|
| 533 |
+
"learning_rate": 5.7035283819124155e-06,
|
| 534 |
+
"loss": 1.0546,
|
| 535 |
+
"step": 325
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 8.986301369863014,
|
| 539 |
+
"eval_loss": 2.5756430625915527,
|
| 540 |
+
"eval_runtime": 0.5591,
|
| 541 |
+
"eval_samples_per_second": 17.886,
|
| 542 |
+
"eval_steps_per_second": 1.789,
|
| 543 |
+
"step": 328
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.04109589041096,
|
| 547 |
+
"grad_norm": 0.33203125,
|
| 548 |
+
"learning_rate": 4.20104876845111e-06,
|
| 549 |
+
"loss": 1.0515,
|
| 550 |
+
"step": 330
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.178082191780822,
|
| 554 |
+
"grad_norm": 0.3515625,
|
| 555 |
+
"learning_rate": 2.9236950338380033e-06,
|
| 556 |
+
"loss": 1.0521,
|
| 557 |
+
"step": 335
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.315068493150685,
|
| 561 |
+
"grad_norm": 0.3359375,
|
| 562 |
+
"learning_rate": 1.874468937261531e-06,
|
| 563 |
+
"loss": 1.0563,
|
| 564 |
+
"step": 340
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.452054794520548,
|
| 568 |
+
"grad_norm": 0.341796875,
|
| 569 |
+
"learning_rate": 1.055836141905553e-06,
|
| 570 |
+
"loss": 1.0535,
|
| 571 |
+
"step": 345
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.58904109589041,
|
| 575 |
+
"grad_norm": 0.349609375,
|
| 576 |
+
"learning_rate": 4.6972042068341714e-07,
|
| 577 |
+
"loss": 1.0521,
|
| 578 |
+
"step": 350
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.726027397260275,
|
| 582 |
+
"grad_norm": 0.341796875,
|
| 583 |
+
"learning_rate": 1.1749913540496371e-07,
|
| 584 |
+
"loss": 1.0664,
|
| 585 |
+
"step": 355
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.863013698630137,
|
| 589 |
+
"grad_norm": 0.345703125,
|
| 590 |
+
"learning_rate": 0.0,
|
| 591 |
+
"loss": 1.0623,
|
| 592 |
+
"step": 360
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.863013698630137,
|
| 596 |
+
"eval_loss": 2.5758092403411865,
|
| 597 |
+
"eval_runtime": 0.5499,
|
| 598 |
+
"eval_samples_per_second": 18.186,
|
| 599 |
+
"eval_steps_per_second": 1.819,
|
| 600 |
+
"step": 360
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 9.863013698630137,
|
| 604 |
+
"step": 360,
|
| 605 |
+
"total_flos": 2.1145440197646746e+17,
|
| 606 |
+
"train_loss": 1.288370986117257,
|
| 607 |
+
"train_runtime": 1930.7273,
|
| 608 |
+
"train_samples_per_second": 9.074,
|
| 609 |
+
"train_steps_per_second": 0.186
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 5,
|
| 613 |
+
"max_steps": 360,
|
| 614 |
+
"num_input_tokens_seen": 0,
|
| 615 |
+
"num_train_epochs": 10,
|
| 616 |
+
"save_steps": 100,
|
| 617 |
+
"stateful_callbacks": {
|
| 618 |
+
"TrainerControl": {
|
| 619 |
+
"args": {
|
| 620 |
+
"should_epoch_stop": false,
|
| 621 |
+
"should_evaluate": false,
|
| 622 |
+
"should_log": false,
|
| 623 |
+
"should_save": true,
|
| 624 |
+
"should_training_stop": false
|
| 625 |
+
},
|
| 626 |
+
"attributes": {}
|
| 627 |
+
}
|
| 628 |
+
},
|
| 629 |
+
"total_flos": 2.1145440197646746e+17,
|
| 630 |
+
"train_batch_size": 8,
|
| 631 |
+
"trial_name": null,
|
| 632 |
+
"trial_params": null
|
| 633 |
+
}
|