Model save
Browse files- README.md +78 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +508 -0
README.md
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
license: llama3.2
|
| 4 |
+
base_model: meta-llama/Llama-3.2-1B
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- sft
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: llama3-1b-summarize-gpt4o-128k
|
| 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 |
+
# llama3-1b-summarize-gpt4o-128k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B](https://huggingface.co/meta-llama/Llama-3.2-1B) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.7256
|
| 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: 32
|
| 44 |
+
- eval_batch_size: 32
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 8
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 512
|
| 50 |
+
- total_eval_batch_size: 256
|
| 51 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 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 |
+
| 2.1182 | 1.0 | 28 | 2.7821 |
|
| 61 |
+
| 1.8354 | 2.0 | 56 | 2.7378 |
|
| 62 |
+
| 1.7604 | 3.0 | 84 | 2.7259 |
|
| 63 |
+
| 1.7035 | 4.0 | 112 | 2.7250 |
|
| 64 |
+
| 1.6822 | 5.0 | 140 | 2.7263 |
|
| 65 |
+
| 1.6619 | 6.0 | 168 | 2.7254 |
|
| 66 |
+
| 1.6471 | 7.0 | 196 | 2.7241 |
|
| 67 |
+
| 1.6458 | 8.0 | 224 | 2.7253 |
|
| 68 |
+
| 1.641 | 9.0 | 252 | 2.7256 |
|
| 69 |
+
| 1.6383 | 9.6545 | 270 | 2.7256 |
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
### Framework versions
|
| 73 |
+
|
| 74 |
+
- PEFT 0.15.1
|
| 75 |
+
- Transformers 4.50.3
|
| 76 |
+
- Pytorch 2.6.0+cu124
|
| 77 |
+
- Datasets 3.5.0
|
| 78 |
+
- Tokenizers 0.21.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.654545454545454,
|
| 3 |
+
"total_flos": 8.156088875152835e+17,
|
| 4 |
+
"train_loss": 1.7710220513520418,
|
| 5 |
+
"train_runtime": 1245.0854,
|
| 6 |
+
"train_samples": 129221,
|
| 7 |
+
"train_samples_per_second": 112.233,
|
| 8 |
+
"train_steps_per_second": 0.217
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.654545454545454,
|
| 3 |
+
"total_flos": 8.156088875152835e+17,
|
| 4 |
+
"train_loss": 1.7710220513520418,
|
| 5 |
+
"train_runtime": 1245.0854,
|
| 6 |
+
"train_samples": 129221,
|
| 7 |
+
"train_samples_per_second": 112.233,
|
| 8 |
+
"train_steps_per_second": 0.217
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,508 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 9.654545454545454,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 270,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03636363636363636,
|
| 14 |
+
"grad_norm": 3.691458225250244,
|
| 15 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 16 |
+
"loss": 2.6618,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.18181818181818182,
|
| 21 |
+
"grad_norm": 3.446274518966675,
|
| 22 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 23 |
+
"loss": 2.6408,
|
| 24 |
+
"step": 5
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.36363636363636365,
|
| 28 |
+
"grad_norm": 2.032472848892212,
|
| 29 |
+
"learning_rate": 7.407407407407407e-05,
|
| 30 |
+
"loss": 2.5317,
|
| 31 |
+
"step": 10
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5454545454545454,
|
| 35 |
+
"grad_norm": 0.9463324546813965,
|
| 36 |
+
"learning_rate": 0.00011111111111111112,
|
| 37 |
+
"loss": 2.36,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7272727272727273,
|
| 42 |
+
"grad_norm": 0.7080094218254089,
|
| 43 |
+
"learning_rate": 0.00014814814814814815,
|
| 44 |
+
"loss": 2.2174,
|
| 45 |
+
"step": 20
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.9090909090909091,
|
| 49 |
+
"grad_norm": 0.5037975311279297,
|
| 50 |
+
"learning_rate": 0.0001851851851851852,
|
| 51 |
+
"loss": 2.1182,
|
| 52 |
+
"step": 25
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_loss": 2.782132625579834,
|
| 57 |
+
"eval_runtime": 0.8292,
|
| 58 |
+
"eval_samples_per_second": 12.06,
|
| 59 |
+
"eval_steps_per_second": 1.206,
|
| 60 |
+
"step": 28
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.0727272727272728,
|
| 64 |
+
"grad_norm": 0.41888633370399475,
|
| 65 |
+
"learning_rate": 0.00019992479525042303,
|
| 66 |
+
"loss": 2.0277,
|
| 67 |
+
"step": 30
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.2545454545454544,
|
| 71 |
+
"grad_norm": 0.27915704250335693,
|
| 72 |
+
"learning_rate": 0.00019946562024066014,
|
| 73 |
+
"loss": 1.9587,
|
| 74 |
+
"step": 35
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.4363636363636363,
|
| 78 |
+
"grad_norm": 0.20056034624576569,
|
| 79 |
+
"learning_rate": 0.00019859096633447965,
|
| 80 |
+
"loss": 1.9087,
|
| 81 |
+
"step": 40
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.6181818181818182,
|
| 85 |
+
"grad_norm": 0.16737522184848785,
|
| 86 |
+
"learning_rate": 0.00019730448705798239,
|
| 87 |
+
"loss": 1.8766,
|
| 88 |
+
"step": 45
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.8,
|
| 92 |
+
"grad_norm": 0.15048423409461975,
|
| 93 |
+
"learning_rate": 0.00019561155617738797,
|
| 94 |
+
"loss": 1.8481,
|
| 95 |
+
"step": 50
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.981818181818182,
|
| 99 |
+
"grad_norm": 0.1224176436662674,
|
| 100 |
+
"learning_rate": 0.000193519245252219,
|
| 101 |
+
"loss": 1.8354,
|
| 102 |
+
"step": 55
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.0,
|
| 106 |
+
"eval_loss": 2.737755537033081,
|
| 107 |
+
"eval_runtime": 0.829,
|
| 108 |
+
"eval_samples_per_second": 12.063,
|
| 109 |
+
"eval_steps_per_second": 1.206,
|
| 110 |
+
"step": 56
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.1454545454545455,
|
| 114 |
+
"grad_norm": 0.1324545294046402,
|
| 115 |
+
"learning_rate": 0.0001910362940966147,
|
| 116 |
+
"loss": 1.8118,
|
| 117 |
+
"step": 60
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 2.327272727272727,
|
| 121 |
+
"grad_norm": 0.11611360311508179,
|
| 122 |
+
"learning_rate": 0.0001881730742721608,
|
| 123 |
+
"loss": 1.7937,
|
| 124 |
+
"step": 65
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.509090909090909,
|
| 128 |
+
"grad_norm": 0.1148991584777832,
|
| 129 |
+
"learning_rate": 0.00018494154576472976,
|
| 130 |
+
"loss": 1.7791,
|
| 131 |
+
"step": 70
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.690909090909091,
|
| 135 |
+
"grad_norm": 0.11438702791929245,
|
| 136 |
+
"learning_rate": 0.00018135520702629675,
|
| 137 |
+
"loss": 1.7654,
|
| 138 |
+
"step": 75
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.8727272727272726,
|
| 142 |
+
"grad_norm": 0.11716635525226593,
|
| 143 |
+
"learning_rate": 0.00017742903859041325,
|
| 144 |
+
"loss": 1.7604,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.0,
|
| 149 |
+
"eval_loss": 2.7259373664855957,
|
| 150 |
+
"eval_runtime": 0.8303,
|
| 151 |
+
"eval_samples_per_second": 12.044,
|
| 152 |
+
"eval_steps_per_second": 1.204,
|
| 153 |
+
"step": 84
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 3.036363636363636,
|
| 157 |
+
"grad_norm": 0.1302882581949234,
|
| 158 |
+
"learning_rate": 0.00017317944049686124,
|
| 159 |
+
"loss": 1.7453,
|
| 160 |
+
"step": 85
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.2181818181818183,
|
| 164 |
+
"grad_norm": 0.12489154189825058,
|
| 165 |
+
"learning_rate": 0.0001686241637868734,
|
| 166 |
+
"loss": 1.7396,
|
| 167 |
+
"step": 90
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 3.4,
|
| 171 |
+
"grad_norm": 0.10804688185453415,
|
| 172 |
+
"learning_rate": 0.0001637822363550706,
|
| 173 |
+
"loss": 1.7272,
|
| 174 |
+
"step": 95
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.581818181818182,
|
| 178 |
+
"grad_norm": 0.1448238343000412,
|
| 179 |
+
"learning_rate": 0.0001586738834678418,
|
| 180 |
+
"loss": 1.7231,
|
| 181 |
+
"step": 100
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 3.7636363636363637,
|
| 185 |
+
"grad_norm": 0.12403673678636551,
|
| 186 |
+
"learning_rate": 0.00015332044328016914,
|
| 187 |
+
"loss": 1.7101,
|
| 188 |
+
"step": 105
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.9454545454545453,
|
| 192 |
+
"grad_norm": 0.11520184576511383,
|
| 193 |
+
"learning_rate": 0.0001477442777037949,
|
| 194 |
+
"loss": 1.7035,
|
| 195 |
+
"step": 110
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.0,
|
| 199 |
+
"eval_loss": 2.724990129470825,
|
| 200 |
+
"eval_runtime": 0.8296,
|
| 201 |
+
"eval_samples_per_second": 12.053,
|
| 202 |
+
"eval_steps_per_second": 1.205,
|
| 203 |
+
"step": 112
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 4.109090909090909,
|
| 207 |
+
"grad_norm": 0.11850611865520477,
|
| 208 |
+
"learning_rate": 0.0001419686789990429,
|
| 209 |
+
"loss": 1.6998,
|
| 210 |
+
"step": 115
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 4.290909090909091,
|
| 214 |
+
"grad_norm": 0.141310453414917,
|
| 215 |
+
"learning_rate": 0.00013601777248047105,
|
| 216 |
+
"loss": 1.6942,
|
| 217 |
+
"step": 120
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 4.472727272727273,
|
| 221 |
+
"grad_norm": 0.14388997852802277,
|
| 222 |
+
"learning_rate": 0.00012991641574276418,
|
| 223 |
+
"loss": 1.6887,
|
| 224 |
+
"step": 125
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 4.654545454545454,
|
| 228 |
+
"grad_norm": 0.11356977373361588,
|
| 229 |
+
"learning_rate": 0.00012369009482781192,
|
| 230 |
+
"loss": 1.6845,
|
| 231 |
+
"step": 130
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 4.836363636363636,
|
| 235 |
+
"grad_norm": 0.13505423069000244,
|
| 236 |
+
"learning_rate": 0.00011736481776669306,
|
| 237 |
+
"loss": 1.6801,
|
| 238 |
+
"step": 135
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 5.0,
|
| 242 |
+
"grad_norm": 0.18071481585502625,
|
| 243 |
+
"learning_rate": 0.00011096700594125318,
|
| 244 |
+
"loss": 1.6822,
|
| 245 |
+
"step": 140
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 5.0,
|
| 249 |
+
"eval_loss": 2.7262730598449707,
|
| 250 |
+
"eval_runtime": 0.8327,
|
| 251 |
+
"eval_samples_per_second": 12.009,
|
| 252 |
+
"eval_steps_per_second": 1.201,
|
| 253 |
+
"step": 140
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 5.181818181818182,
|
| 257 |
+
"grad_norm": 0.12405228614807129,
|
| 258 |
+
"learning_rate": 0.00010452338371907064,
|
| 259 |
+
"loss": 1.671,
|
| 260 |
+
"step": 145
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 5.363636363636363,
|
| 264 |
+
"grad_norm": 0.15709254145622253,
|
| 265 |
+
"learning_rate": 9.806086682281758e-05,
|
| 266 |
+
"loss": 1.6697,
|
| 267 |
+
"step": 150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 5.545454545454545,
|
| 271 |
+
"grad_norm": 0.1405353993177414,
|
| 272 |
+
"learning_rate": 9.160644990030931e-05,
|
| 273 |
+
"loss": 1.6707,
|
| 274 |
+
"step": 155
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 5.7272727272727275,
|
| 278 |
+
"grad_norm": 0.13487176597118378,
|
| 279 |
+
"learning_rate": 8.518709376487515e-05,
|
| 280 |
+
"loss": 1.6619,
|
| 281 |
+
"step": 160
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 5.909090909090909,
|
| 285 |
+
"grad_norm": 0.12394227087497711,
|
| 286 |
+
"learning_rate": 7.882961277705895e-05,
|
| 287 |
+
"loss": 1.6619,
|
| 288 |
+
"step": 165
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 6.0,
|
| 292 |
+
"eval_loss": 2.7253997325897217,
|
| 293 |
+
"eval_runtime": 0.8321,
|
| 294 |
+
"eval_samples_per_second": 12.017,
|
| 295 |
+
"eval_steps_per_second": 1.202,
|
| 296 |
+
"step": 168
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 6.072727272727272,
|
| 300 |
+
"grad_norm": 0.11816684156656265,
|
| 301 |
+
"learning_rate": 7.256056283806986e-05,
|
| 302 |
+
"loss": 1.6573,
|
| 303 |
+
"step": 170
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.254545454545455,
|
| 307 |
+
"grad_norm": 0.14117498695850372,
|
| 308 |
+
"learning_rate": 6.640613046284581e-05,
|
| 309 |
+
"loss": 1.6622,
|
| 310 |
+
"step": 175
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.4363636363636365,
|
| 314 |
+
"grad_norm": 0.1342514008283615,
|
| 315 |
+
"learning_rate": 6.039202339608432e-05,
|
| 316 |
+
"loss": 1.6535,
|
| 317 |
+
"step": 180
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.618181818181818,
|
| 321 |
+
"grad_norm": 0.13483189046382904,
|
| 322 |
+
"learning_rate": 5.4543363228149946e-05,
|
| 323 |
+
"loss": 1.6532,
|
| 324 |
+
"step": 185
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.8,
|
| 328 |
+
"grad_norm": 0.1636153757572174,
|
| 329 |
+
"learning_rate": 4.888458045941269e-05,
|
| 330 |
+
"loss": 1.6482,
|
| 331 |
+
"step": 190
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.9818181818181815,
|
| 335 |
+
"grad_norm": 0.1563912183046341,
|
| 336 |
+
"learning_rate": 4.343931245134616e-05,
|
| 337 |
+
"loss": 1.6471,
|
| 338 |
+
"step": 195
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.0,
|
| 342 |
+
"eval_loss": 2.7240517139434814,
|
| 343 |
+
"eval_runtime": 0.8312,
|
| 344 |
+
"eval_samples_per_second": 12.031,
|
| 345 |
+
"eval_steps_per_second": 1.203,
|
| 346 |
+
"step": 196
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 7.1454545454545455,
|
| 350 |
+
"grad_norm": 0.11320989578962326,
|
| 351 |
+
"learning_rate": 3.8230304690654304e-05,
|
| 352 |
+
"loss": 1.6472,
|
| 353 |
+
"step": 200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 7.327272727272727,
|
| 357 |
+
"grad_norm": 0.111383818089962,
|
| 358 |
+
"learning_rate": 3.3279315778858036e-05,
|
| 359 |
+
"loss": 1.6488,
|
| 360 |
+
"step": 205
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 7.509090909090909,
|
| 364 |
+
"grad_norm": 0.10844731330871582,
|
| 365 |
+
"learning_rate": 2.8607026544210114e-05,
|
| 366 |
+
"loss": 1.6458,
|
| 367 |
+
"step": 210
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.690909090909091,
|
| 371 |
+
"grad_norm": 0.10823339223861694,
|
| 372 |
+
"learning_rate": 2.423295365558821e-05,
|
| 373 |
+
"loss": 1.6456,
|
| 374 |
+
"step": 215
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 7.872727272727273,
|
| 378 |
+
"grad_norm": 0.10790830850601196,
|
| 379 |
+
"learning_rate": 2.01753680992107e-05,
|
| 380 |
+
"loss": 1.6458,
|
| 381 |
+
"step": 220
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.0,
|
| 385 |
+
"eval_loss": 2.7252650260925293,
|
| 386 |
+
"eval_runtime": 0.8302,
|
| 387 |
+
"eval_samples_per_second": 12.045,
|
| 388 |
+
"eval_steps_per_second": 1.204,
|
| 389 |
+
"step": 224
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 8.036363636363637,
|
| 393 |
+
"grad_norm": 0.11462420970201492,
|
| 394 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 395 |
+
"loss": 1.643,
|
| 396 |
+
"step": 225
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 8.218181818181819,
|
| 400 |
+
"grad_norm": 0.10164881497621536,
|
| 401 |
+
"learning_rate": 1.307606211733522e-05,
|
| 402 |
+
"loss": 1.6435,
|
| 403 |
+
"step": 230
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 8.4,
|
| 407 |
+
"grad_norm": 0.11715802550315857,
|
| 408 |
+
"learning_rate": 1.0063996278090704e-05,
|
| 409 |
+
"loss": 1.6436,
|
| 410 |
+
"step": 235
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 8.581818181818182,
|
| 414 |
+
"grad_norm": 0.1077931597828865,
|
| 415 |
+
"learning_rate": 7.427603073110967e-06,
|
| 416 |
+
"loss": 1.6437,
|
| 417 |
+
"step": 240
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 8.763636363636364,
|
| 421 |
+
"grad_norm": 0.09881118685007095,
|
| 422 |
+
"learning_rate": 5.177895008392353e-06,
|
| 423 |
+
"loss": 1.6415,
|
| 424 |
+
"step": 245
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 8.945454545454545,
|
| 428 |
+
"grad_norm": 0.0973580852150917,
|
| 429 |
+
"learning_rate": 3.3242693633337983e-06,
|
| 430 |
+
"loss": 1.641,
|
| 431 |
+
"step": 250
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 9.0,
|
| 435 |
+
"eval_loss": 2.725569009780884,
|
| 436 |
+
"eval_runtime": 0.8306,
|
| 437 |
+
"eval_samples_per_second": 12.039,
|
| 438 |
+
"eval_steps_per_second": 1.204,
|
| 439 |
+
"step": 252
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 9.10909090909091,
|
| 443 |
+
"grad_norm": 0.10264136642217636,
|
| 444 |
+
"learning_rate": 1.874468937261531e-06,
|
| 445 |
+
"loss": 1.6464,
|
| 446 |
+
"step": 255
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 9.290909090909091,
|
| 450 |
+
"grad_norm": 0.1021399274468422,
|
| 451 |
+
"learning_rate": 8.345497068998897e-07,
|
| 452 |
+
"loss": 1.6443,
|
| 453 |
+
"step": 260
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 9.472727272727273,
|
| 457 |
+
"grad_norm": 0.10423731058835983,
|
| 458 |
+
"learning_rate": 2.088555298867978e-07,
|
| 459 |
+
"loss": 1.6436,
|
| 460 |
+
"step": 265
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 9.654545454545454,
|
| 464 |
+
"grad_norm": 0.09860274940729141,
|
| 465 |
+
"learning_rate": 0.0,
|
| 466 |
+
"loss": 1.6383,
|
| 467 |
+
"step": 270
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 9.654545454545454,
|
| 471 |
+
"eval_loss": 2.725593328475952,
|
| 472 |
+
"eval_runtime": 0.8317,
|
| 473 |
+
"eval_samples_per_second": 12.024,
|
| 474 |
+
"eval_steps_per_second": 1.202,
|
| 475 |
+
"step": 270
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 9.654545454545454,
|
| 479 |
+
"step": 270,
|
| 480 |
+
"total_flos": 8.156088875152835e+17,
|
| 481 |
+
"train_loss": 1.7710220513520418,
|
| 482 |
+
"train_runtime": 1245.0854,
|
| 483 |
+
"train_samples_per_second": 112.233,
|
| 484 |
+
"train_steps_per_second": 0.217
|
| 485 |
+
}
|
| 486 |
+
],
|
| 487 |
+
"logging_steps": 5,
|
| 488 |
+
"max_steps": 270,
|
| 489 |
+
"num_input_tokens_seen": 0,
|
| 490 |
+
"num_train_epochs": 10,
|
| 491 |
+
"save_steps": 100,
|
| 492 |
+
"stateful_callbacks": {
|
| 493 |
+
"TrainerControl": {
|
| 494 |
+
"args": {
|
| 495 |
+
"should_epoch_stop": false,
|
| 496 |
+
"should_evaluate": false,
|
| 497 |
+
"should_log": false,
|
| 498 |
+
"should_save": true,
|
| 499 |
+
"should_training_stop": true
|
| 500 |
+
},
|
| 501 |
+
"attributes": {}
|
| 502 |
+
}
|
| 503 |
+
},
|
| 504 |
+
"total_flos": 8.156088875152835e+17,
|
| 505 |
+
"train_batch_size": 32,
|
| 506 |
+
"trial_name": null,
|
| 507 |
+
"trial_params": null
|
| 508 |
+
}
|