Model save
Browse files- README.md +78 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +578 -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-closedqa-gpt4o-100k
|
| 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-closedqa-gpt4o-100k
|
| 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: 2.5954
|
| 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 |
+
| 1.845 | 1.0 | 32 | 2.6849 |
|
| 61 |
+
| 1.6455 | 2.0 | 64 | 2.6129 |
|
| 62 |
+
| 1.5798 | 3.0 | 96 | 2.6014 |
|
| 63 |
+
| 1.546 | 4.0 | 128 | 2.5975 |
|
| 64 |
+
| 1.5244 | 5.0 | 160 | 2.5963 |
|
| 65 |
+
| 1.5072 | 6.0 | 192 | 2.5944 |
|
| 66 |
+
| 1.501 | 7.0 | 224 | 2.5941 |
|
| 67 |
+
| 1.4858 | 8.0 | 256 | 2.5944 |
|
| 68 |
+
| 1.4917 | 9.0 | 288 | 2.5956 |
|
| 69 |
+
| 1.4886 | 10.0 | 320 | 2.5954 |
|
| 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": 10.0,
|
| 3 |
+
"total_flos": 9.830314275137126e+17,
|
| 4 |
+
"train_loss": 1.5905956603586673,
|
| 5 |
+
"train_runtime": 1497.0202,
|
| 6 |
+
"train_samples": 111440,
|
| 7 |
+
"train_samples_per_second": 109.25,
|
| 8 |
+
"train_steps_per_second": 0.214
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 9.830314275137126e+17,
|
| 4 |
+
"train_loss": 1.5905956603586673,
|
| 5 |
+
"train_runtime": 1497.0202,
|
| 6 |
+
"train_samples": 111440,
|
| 7 |
+
"train_samples_per_second": 109.25,
|
| 8 |
+
"train_steps_per_second": 0.214
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,578 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 320,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03125,
|
| 14 |
+
"grad_norm": 2.2531771659851074,
|
| 15 |
+
"learning_rate": 6.25e-06,
|
| 16 |
+
"loss": 2.2918,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.15625,
|
| 21 |
+
"grad_norm": 2.1257269382476807,
|
| 22 |
+
"learning_rate": 3.125e-05,
|
| 23 |
+
"loss": 2.284,
|
| 24 |
+
"step": 5
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3125,
|
| 28 |
+
"grad_norm": 1.5896614789962769,
|
| 29 |
+
"learning_rate": 6.25e-05,
|
| 30 |
+
"loss": 2.2225,
|
| 31 |
+
"step": 10
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.46875,
|
| 35 |
+
"grad_norm": 0.7286556363105774,
|
| 36 |
+
"learning_rate": 9.375e-05,
|
| 37 |
+
"loss": 2.1021,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.625,
|
| 42 |
+
"grad_norm": 0.6398385763168335,
|
| 43 |
+
"learning_rate": 0.000125,
|
| 44 |
+
"loss": 1.9996,
|
| 45 |
+
"step": 20
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.78125,
|
| 49 |
+
"grad_norm": 0.4788893461227417,
|
| 50 |
+
"learning_rate": 0.00015625,
|
| 51 |
+
"loss": 1.9178,
|
| 52 |
+
"step": 25
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9375,
|
| 56 |
+
"grad_norm": 0.3934371769428253,
|
| 57 |
+
"learning_rate": 0.0001875,
|
| 58 |
+
"loss": 1.845,
|
| 59 |
+
"step": 30
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_loss": 2.684882402420044,
|
| 64 |
+
"eval_runtime": 0.8301,
|
| 65 |
+
"eval_samples_per_second": 7.228,
|
| 66 |
+
"eval_steps_per_second": 1.205,
|
| 67 |
+
"step": 32
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.09375,
|
| 71 |
+
"grad_norm": 0.28933560848236084,
|
| 72 |
+
"learning_rate": 0.00019994645874763658,
|
| 73 |
+
"loss": 1.786,
|
| 74 |
+
"step": 35
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.25,
|
| 78 |
+
"grad_norm": 0.20734435319900513,
|
| 79 |
+
"learning_rate": 0.00019961946980917456,
|
| 80 |
+
"loss": 1.7402,
|
| 81 |
+
"step": 40
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.40625,
|
| 85 |
+
"grad_norm": 0.16501550376415253,
|
| 86 |
+
"learning_rate": 0.00019899620837148077,
|
| 87 |
+
"loss": 1.7072,
|
| 88 |
+
"step": 45
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.5625,
|
| 92 |
+
"grad_norm": 0.13184033334255219,
|
| 93 |
+
"learning_rate": 0.00019807852804032305,
|
| 94 |
+
"loss": 1.6846,
|
| 95 |
+
"step": 50
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.71875,
|
| 99 |
+
"grad_norm": 0.11590086668729782,
|
| 100 |
+
"learning_rate": 0.00019686915803565934,
|
| 101 |
+
"loss": 1.665,
|
| 102 |
+
"step": 55
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.875,
|
| 106 |
+
"grad_norm": 0.11442063748836517,
|
| 107 |
+
"learning_rate": 0.0001953716950748227,
|
| 108 |
+
"loss": 1.6455,
|
| 109 |
+
"step": 60
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.0,
|
| 113 |
+
"eval_loss": 2.612865686416626,
|
| 114 |
+
"eval_runtime": 0.83,
|
| 115 |
+
"eval_samples_per_second": 7.229,
|
| 116 |
+
"eval_steps_per_second": 1.205,
|
| 117 |
+
"step": 64
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 2.03125,
|
| 121 |
+
"grad_norm": 0.09601253271102905,
|
| 122 |
+
"learning_rate": 0.0001935905926757326,
|
| 123 |
+
"loss": 1.6364,
|
| 124 |
+
"step": 65
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.1875,
|
| 128 |
+
"grad_norm": 0.10327372699975967,
|
| 129 |
+
"learning_rate": 0.00019153114791194473,
|
| 130 |
+
"loss": 1.6227,
|
| 131 |
+
"step": 70
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.34375,
|
| 135 |
+
"grad_norm": 0.09720147401094437,
|
| 136 |
+
"learning_rate": 0.00018919948565893142,
|
| 137 |
+
"loss": 1.611,
|
| 138 |
+
"step": 75
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.5,
|
| 142 |
+
"grad_norm": 0.09051357209682465,
|
| 143 |
+
"learning_rate": 0.00018660254037844388,
|
| 144 |
+
"loss": 1.6067,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 2.65625,
|
| 149 |
+
"grad_norm": 0.10204631090164185,
|
| 150 |
+
"learning_rate": 0.0001837480354951308,
|
| 151 |
+
"loss": 1.6009,
|
| 152 |
+
"step": 85
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.8125,
|
| 156 |
+
"grad_norm": 0.11083294451236725,
|
| 157 |
+
"learning_rate": 0.00018064446042674828,
|
| 158 |
+
"loss": 1.5825,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.96875,
|
| 163 |
+
"grad_norm": 0.1064315065741539,
|
| 164 |
+
"learning_rate": 0.0001773010453362737,
|
| 165 |
+
"loss": 1.5798,
|
| 166 |
+
"step": 95
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 3.0,
|
| 170 |
+
"eval_loss": 2.6013729572296143,
|
| 171 |
+
"eval_runtime": 0.8292,
|
| 172 |
+
"eval_samples_per_second": 7.236,
|
| 173 |
+
"eval_steps_per_second": 1.206,
|
| 174 |
+
"step": 96
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.125,
|
| 178 |
+
"grad_norm": 0.0996951013803482,
|
| 179 |
+
"learning_rate": 0.0001737277336810124,
|
| 180 |
+
"loss": 1.5716,
|
| 181 |
+
"step": 100
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 3.28125,
|
| 185 |
+
"grad_norm": 0.10399119555950165,
|
| 186 |
+
"learning_rate": 0.00016993515264033672,
|
| 187 |
+
"loss": 1.5691,
|
| 188 |
+
"step": 105
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.4375,
|
| 192 |
+
"grad_norm": 0.0989290326833725,
|
| 193 |
+
"learning_rate": 0.00016593458151000688,
|
| 194 |
+
"loss": 1.5641,
|
| 195 |
+
"step": 110
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.59375,
|
| 199 |
+
"grad_norm": 0.108305424451828,
|
| 200 |
+
"learning_rate": 0.00016173791815707051,
|
| 201 |
+
"loss": 1.5539,
|
| 202 |
+
"step": 115
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 3.75,
|
| 206 |
+
"grad_norm": 0.11103138327598572,
|
| 207 |
+
"learning_rate": 0.0001573576436351046,
|
| 208 |
+
"loss": 1.5511,
|
| 209 |
+
"step": 120
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 3.90625,
|
| 213 |
+
"grad_norm": 0.11553769558668137,
|
| 214 |
+
"learning_rate": 0.0001528067850650368,
|
| 215 |
+
"loss": 1.546,
|
| 216 |
+
"step": 125
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 4.0,
|
| 220 |
+
"eval_loss": 2.597513437271118,
|
| 221 |
+
"eval_runtime": 0.8293,
|
| 222 |
+
"eval_samples_per_second": 7.235,
|
| 223 |
+
"eval_steps_per_second": 1.206,
|
| 224 |
+
"step": 128
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 4.0625,
|
| 228 |
+
"grad_norm": 0.11389793455600739,
|
| 229 |
+
"learning_rate": 0.00014809887689193877,
|
| 230 |
+
"loss": 1.5361,
|
| 231 |
+
"step": 130
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 4.21875,
|
| 235 |
+
"grad_norm": 0.1214594691991806,
|
| 236 |
+
"learning_rate": 0.00014324792063301662,
|
| 237 |
+
"loss": 1.5346,
|
| 238 |
+
"step": 135
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 4.375,
|
| 242 |
+
"grad_norm": 0.11472882330417633,
|
| 243 |
+
"learning_rate": 0.000138268343236509,
|
| 244 |
+
"loss": 1.5339,
|
| 245 |
+
"step": 140
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 4.53125,
|
| 249 |
+
"grad_norm": 0.10563024878501892,
|
| 250 |
+
"learning_rate": 0.00013317495417533524,
|
| 251 |
+
"loss": 1.5314,
|
| 252 |
+
"step": 145
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.6875,
|
| 256 |
+
"grad_norm": 0.10731685161590576,
|
| 257 |
+
"learning_rate": 0.00012798290140309923,
|
| 258 |
+
"loss": 1.5284,
|
| 259 |
+
"step": 150
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 4.84375,
|
| 263 |
+
"grad_norm": 0.10968785732984543,
|
| 264 |
+
"learning_rate": 0.00012270762630343734,
|
| 265 |
+
"loss": 1.5233,
|
| 266 |
+
"step": 155
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 5.0,
|
| 270 |
+
"grad_norm": 0.11469796299934387,
|
| 271 |
+
"learning_rate": 0.00011736481776669306,
|
| 272 |
+
"loss": 1.5244,
|
| 273 |
+
"step": 160
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 5.0,
|
| 277 |
+
"eval_loss": 2.5962610244750977,
|
| 278 |
+
"eval_runtime": 0.8296,
|
| 279 |
+
"eval_samples_per_second": 7.233,
|
| 280 |
+
"eval_steps_per_second": 1.205,
|
| 281 |
+
"step": 160
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 5.15625,
|
| 285 |
+
"grad_norm": 0.10656954348087311,
|
| 286 |
+
"learning_rate": 0.00011197036553049625,
|
| 287 |
+
"loss": 1.5151,
|
| 288 |
+
"step": 165
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 5.3125,
|
| 292 |
+
"grad_norm": 0.12216157466173172,
|
| 293 |
+
"learning_rate": 0.00010654031292301432,
|
| 294 |
+
"loss": 1.5119,
|
| 295 |
+
"step": 170
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 5.46875,
|
| 299 |
+
"grad_norm": 0.11443573981523514,
|
| 300 |
+
"learning_rate": 0.00010109080914941824,
|
| 301 |
+
"loss": 1.5107,
|
| 302 |
+
"step": 175
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 5.625,
|
| 306 |
+
"grad_norm": 0.11616319417953491,
|
| 307 |
+
"learning_rate": 9.563806126346642e-05,
|
| 308 |
+
"loss": 1.5165,
|
| 309 |
+
"step": 180
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.78125,
|
| 313 |
+
"grad_norm": 0.1112031489610672,
|
| 314 |
+
"learning_rate": 9.019828596704394e-05,
|
| 315 |
+
"loss": 1.509,
|
| 316 |
+
"step": 185
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.9375,
|
| 320 |
+
"grad_norm": 0.11148490756750107,
|
| 321 |
+
"learning_rate": 8.478766138100834e-05,
|
| 322 |
+
"loss": 1.5072,
|
| 323 |
+
"step": 190
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 6.0,
|
| 327 |
+
"eval_loss": 2.594440460205078,
|
| 328 |
+
"eval_runtime": 0.8285,
|
| 329 |
+
"eval_samples_per_second": 7.242,
|
| 330 |
+
"eval_steps_per_second": 1.207,
|
| 331 |
+
"step": 192
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.09375,
|
| 335 |
+
"grad_norm": 0.11521880328655243,
|
| 336 |
+
"learning_rate": 7.942227893077652e-05,
|
| 337 |
+
"loss": 1.5015,
|
| 338 |
+
"step": 195
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.25,
|
| 342 |
+
"grad_norm": 0.11160895973443985,
|
| 343 |
+
"learning_rate": 7.411809548974792e-05,
|
| 344 |
+
"loss": 1.4991,
|
| 345 |
+
"step": 200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.40625,
|
| 349 |
+
"grad_norm": 0.11868108808994293,
|
| 350 |
+
"learning_rate": 6.889088592289093e-05,
|
| 351 |
+
"loss": 1.5008,
|
| 352 |
+
"step": 205
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.5625,
|
| 356 |
+
"grad_norm": 0.11169662326574326,
|
| 357 |
+
"learning_rate": 6.375619617162985e-05,
|
| 358 |
+
"loss": 1.5006,
|
| 359 |
+
"step": 210
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.71875,
|
| 363 |
+
"grad_norm": 0.11667145788669586,
|
| 364 |
+
"learning_rate": 5.872929701956054e-05,
|
| 365 |
+
"loss": 1.4989,
|
| 366 |
+
"step": 215
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.875,
|
| 370 |
+
"grad_norm": 0.10703334957361221,
|
| 371 |
+
"learning_rate": 5.382513867649663e-05,
|
| 372 |
+
"loss": 1.501,
|
| 373 |
+
"step": 220
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.0,
|
| 377 |
+
"eval_loss": 2.594101667404175,
|
| 378 |
+
"eval_runtime": 0.8304,
|
| 379 |
+
"eval_samples_per_second": 7.225,
|
| 380 |
+
"eval_steps_per_second": 1.204,
|
| 381 |
+
"step": 224
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 7.03125,
|
| 385 |
+
"grad_norm": 0.11598368734121323,
|
| 386 |
+
"learning_rate": 4.9058306315915826e-05,
|
| 387 |
+
"loss": 1.5005,
|
| 388 |
+
"step": 225
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 7.1875,
|
| 392 |
+
"grad_norm": 0.11565965414047241,
|
| 393 |
+
"learning_rate": 4.444297669803981e-05,
|
| 394 |
+
"loss": 1.4975,
|
| 395 |
+
"step": 230
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 7.34375,
|
| 399 |
+
"grad_norm": 0.10369732230901718,
|
| 400 |
+
"learning_rate": 3.999287600755192e-05,
|
| 401 |
+
"loss": 1.4969,
|
| 402 |
+
"step": 235
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.5,
|
| 406 |
+
"grad_norm": 0.10466954857110977,
|
| 407 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 408 |
+
"loss": 1.4949,
|
| 409 |
+
"step": 240
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 7.65625,
|
| 413 |
+
"grad_norm": 0.09984403848648071,
|
| 414 |
+
"learning_rate": 3.164076979771287e-05,
|
| 415 |
+
"loss": 1.4901,
|
| 416 |
+
"step": 245
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.8125,
|
| 420 |
+
"grad_norm": 0.10001492500305176,
|
| 421 |
+
"learning_rate": 2.776360379402445e-05,
|
| 422 |
+
"loss": 1.492,
|
| 423 |
+
"step": 250
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.96875,
|
| 427 |
+
"grad_norm": 0.10096081346273422,
|
| 428 |
+
"learning_rate": 2.4101271875283817e-05,
|
| 429 |
+
"loss": 1.4858,
|
| 430 |
+
"step": 255
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 8.0,
|
| 434 |
+
"eval_loss": 2.5943994522094727,
|
| 435 |
+
"eval_runtime": 0.8348,
|
| 436 |
+
"eval_samples_per_second": 7.187,
|
| 437 |
+
"eval_steps_per_second": 1.198,
|
| 438 |
+
"step": 256
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.125,
|
| 442 |
+
"grad_norm": 0.09807378798723221,
|
| 443 |
+
"learning_rate": 2.0664665970876496e-05,
|
| 444 |
+
"loss": 1.4889,
|
| 445 |
+
"step": 260
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 8.28125,
|
| 449 |
+
"grad_norm": 0.1029752716422081,
|
| 450 |
+
"learning_rate": 1.7464006691513623e-05,
|
| 451 |
+
"loss": 1.4893,
|
| 452 |
+
"step": 265
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 8.4375,
|
| 456 |
+
"grad_norm": 0.0935431718826294,
|
| 457 |
+
"learning_rate": 1.4508812932705363e-05,
|
| 458 |
+
"loss": 1.4883,
|
| 459 |
+
"step": 270
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 8.59375,
|
| 463 |
+
"grad_norm": 0.09739290177822113,
|
| 464 |
+
"learning_rate": 1.1807873565164506e-05,
|
| 465 |
+
"loss": 1.4858,
|
| 466 |
+
"step": 275
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.75,
|
| 470 |
+
"grad_norm": 0.10472334176301956,
|
| 471 |
+
"learning_rate": 9.369221296335006e-06,
|
| 472 |
+
"loss": 1.4933,
|
| 473 |
+
"step": 280
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 8.90625,
|
| 477 |
+
"grad_norm": 0.09686731547117233,
|
| 478 |
+
"learning_rate": 7.200108780781556e-06,
|
| 479 |
+
"loss": 1.4917,
|
| 480 |
+
"step": 285
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 9.0,
|
| 484 |
+
"eval_loss": 2.5955541133880615,
|
| 485 |
+
"eval_runtime": 0.8323,
|
| 486 |
+
"eval_samples_per_second": 7.209,
|
| 487 |
+
"eval_steps_per_second": 1.201,
|
| 488 |
+
"step": 288
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 9.0625,
|
| 492 |
+
"grad_norm": 0.09799981117248535,
|
| 493 |
+
"learning_rate": 5.306987050489442e-06,
|
| 494 |
+
"loss": 1.4904,
|
| 495 |
+
"step": 290
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 9.21875,
|
| 499 |
+
"grad_norm": 0.10055091232061386,
|
| 500 |
+
"learning_rate": 3.6954863292237297e-06,
|
| 501 |
+
"loss": 1.4813,
|
| 502 |
+
"step": 295
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 9.375,
|
| 506 |
+
"grad_norm": 0.09550856798887253,
|
| 507 |
+
"learning_rate": 2.3703992880066638e-06,
|
| 508 |
+
"loss": 1.4873,
|
| 509 |
+
"step": 300
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 9.53125,
|
| 513 |
+
"grad_norm": 0.09861158579587936,
|
| 514 |
+
"learning_rate": 1.3356667915121025e-06,
|
| 515 |
+
"loss": 1.4911,
|
| 516 |
+
"step": 305
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 9.6875,
|
| 520 |
+
"grad_norm": 0.10015583783388138,
|
| 521 |
+
"learning_rate": 5.943661777680354e-07,
|
| 522 |
+
"loss": 1.4933,
|
| 523 |
+
"step": 310
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 9.84375,
|
| 527 |
+
"grad_norm": 0.09337816387414932,
|
| 528 |
+
"learning_rate": 1.487021060236904e-07,
|
| 529 |
+
"loss": 1.49,
|
| 530 |
+
"step": 315
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 10.0,
|
| 534 |
+
"grad_norm": 0.08882776647806168,
|
| 535 |
+
"learning_rate": 0.0,
|
| 536 |
+
"loss": 1.4886,
|
| 537 |
+
"step": 320
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 10.0,
|
| 541 |
+
"eval_loss": 2.595360517501831,
|
| 542 |
+
"eval_runtime": 0.827,
|
| 543 |
+
"eval_samples_per_second": 7.255,
|
| 544 |
+
"eval_steps_per_second": 1.209,
|
| 545 |
+
"step": 320
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 10.0,
|
| 549 |
+
"step": 320,
|
| 550 |
+
"total_flos": 9.830314275137126e+17,
|
| 551 |
+
"train_loss": 1.5905956603586673,
|
| 552 |
+
"train_runtime": 1497.0202,
|
| 553 |
+
"train_samples_per_second": 109.25,
|
| 554 |
+
"train_steps_per_second": 0.214
|
| 555 |
+
}
|
| 556 |
+
],
|
| 557 |
+
"logging_steps": 5,
|
| 558 |
+
"max_steps": 320,
|
| 559 |
+
"num_input_tokens_seen": 0,
|
| 560 |
+
"num_train_epochs": 10,
|
| 561 |
+
"save_steps": 100,
|
| 562 |
+
"stateful_callbacks": {
|
| 563 |
+
"TrainerControl": {
|
| 564 |
+
"args": {
|
| 565 |
+
"should_epoch_stop": false,
|
| 566 |
+
"should_evaluate": false,
|
| 567 |
+
"should_log": false,
|
| 568 |
+
"should_save": true,
|
| 569 |
+
"should_training_stop": true
|
| 570 |
+
},
|
| 571 |
+
"attributes": {}
|
| 572 |
+
}
|
| 573 |
+
},
|
| 574 |
+
"total_flos": 9.830314275137126e+17,
|
| 575 |
+
"train_batch_size": 32,
|
| 576 |
+
"trial_name": null,
|
| 577 |
+
"trial_params": null
|
| 578 |
+
}
|