Model save
Browse files- README.md +78 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +606 -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: llama3.1-1b-coding-gpt4o-100k2
|
| 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 |
+
# llama3.1-1b-coding-gpt4o-100k2
|
| 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: 1.6745
|
| 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: 0.002
|
| 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 |
+
| 1.2024 | 1.0 | 34 | 1.7381 |
|
| 61 |
+
| 1.0846 | 2.0 | 68 | 1.6923 |
|
| 62 |
+
| 1.0447 | 3.0 | 102 | 1.6731 |
|
| 63 |
+
| 1.0207 | 4.0 | 136 | 1.6660 |
|
| 64 |
+
| 1.0039 | 5.0 | 170 | 1.6681 |
|
| 65 |
+
| 0.9957 | 6.0 | 204 | 1.6620 |
|
| 66 |
+
| 0.9793 | 7.0 | 238 | 1.6656 |
|
| 67 |
+
| 0.9761 | 8.0 | 272 | 1.6707 |
|
| 68 |
+
| 0.9678 | 9.0 | 306 | 1.6741 |
|
| 69 |
+
| 0.9709 | 10.0 | 340 | 1.6745 |
|
| 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": 1.0444708917333197e+18,
|
| 4 |
+
"train_loss": 1.0473914388348073,
|
| 5 |
+
"train_runtime": 1591.5084,
|
| 6 |
+
"train_samples": 116368,
|
| 7 |
+
"train_samples_per_second": 108.407,
|
| 8 |
+
"train_steps_per_second": 0.214
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 1.0444708917333197e+18,
|
| 4 |
+
"train_loss": 1.0473914388348073,
|
| 5 |
+
"train_runtime": 1591.5084,
|
| 6 |
+
"train_samples": 116368,
|
| 7 |
+
"train_samples_per_second": 108.407,
|
| 8 |
+
"train_steps_per_second": 0.214
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,606 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 340,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.029411764705882353,
|
| 14 |
+
"grad_norm": 2.488327741622925,
|
| 15 |
+
"learning_rate": 5.882352941176471e-05,
|
| 16 |
+
"loss": 1.9016,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.14705882352941177,
|
| 21 |
+
"grad_norm": 1.427666425704956,
|
| 22 |
+
"learning_rate": 0.00029411764705882356,
|
| 23 |
+
"loss": 1.8391,
|
| 24 |
+
"step": 5
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.29411764705882354,
|
| 28 |
+
"grad_norm": 0.6930230259895325,
|
| 29 |
+
"learning_rate": 0.0005882352941176471,
|
| 30 |
+
"loss": 1.5625,
|
| 31 |
+
"step": 10
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.4411764705882353,
|
| 35 |
+
"grad_norm": 0.2596357762813568,
|
| 36 |
+
"learning_rate": 0.0008823529411764706,
|
| 37 |
+
"loss": 1.3849,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.5882352941176471,
|
| 42 |
+
"grad_norm": 0.16956600546836853,
|
| 43 |
+
"learning_rate": 0.0011764705882352942,
|
| 44 |
+
"loss": 1.2969,
|
| 45 |
+
"step": 20
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7352941176470589,
|
| 49 |
+
"grad_norm": 0.12523840367794037,
|
| 50 |
+
"learning_rate": 0.0014705882352941178,
|
| 51 |
+
"loss": 1.2405,
|
| 52 |
+
"step": 25
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.8823529411764706,
|
| 56 |
+
"grad_norm": 0.09968027472496033,
|
| 57 |
+
"learning_rate": 0.0017647058823529412,
|
| 58 |
+
"loss": 1.2024,
|
| 59 |
+
"step": 30
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_loss": 1.7381153106689453,
|
| 64 |
+
"eval_runtime": 0.828,
|
| 65 |
+
"eval_samples_per_second": 4.831,
|
| 66 |
+
"eval_steps_per_second": 1.208,
|
| 67 |
+
"step": 34
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.0294117647058822,
|
| 71 |
+
"grad_norm": 0.366025447845459,
|
| 72 |
+
"learning_rate": 0.001999947298487173,
|
| 73 |
+
"loss": 1.1724,
|
| 74 |
+
"step": 35
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.1764705882352942,
|
| 78 |
+
"grad_norm": 0.12924526631832123,
|
| 79 |
+
"learning_rate": 0.0019981033287370443,
|
| 80 |
+
"loss": 1.1507,
|
| 81 |
+
"step": 40
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.3235294117647058,
|
| 85 |
+
"grad_norm": 0.15898454189300537,
|
| 86 |
+
"learning_rate": 0.0019936298356132177,
|
| 87 |
+
"loss": 1.1306,
|
| 88 |
+
"step": 45
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.4705882352941178,
|
| 92 |
+
"grad_norm": 0.09307724982500076,
|
| 93 |
+
"learning_rate": 0.0019865386046236597,
|
| 94 |
+
"loss": 1.1156,
|
| 95 |
+
"step": 50
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.6176470588235294,
|
| 99 |
+
"grad_norm": 0.0881420224905014,
|
| 100 |
+
"learning_rate": 0.001976848317759601,
|
| 101 |
+
"loss": 1.1038,
|
| 102 |
+
"step": 55
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.7647058823529411,
|
| 106 |
+
"grad_norm": 0.08049994707107544,
|
| 107 |
+
"learning_rate": 0.0019645845042774554,
|
| 108 |
+
"loss": 1.0939,
|
| 109 |
+
"step": 60
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.9117647058823528,
|
| 113 |
+
"grad_norm": 0.07463372498750687,
|
| 114 |
+
"learning_rate": 0.001949779473441478,
|
| 115 |
+
"loss": 1.0846,
|
| 116 |
+
"step": 65
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.0,
|
| 120 |
+
"eval_loss": 1.6922645568847656,
|
| 121 |
+
"eval_runtime": 0.8305,
|
| 122 |
+
"eval_samples_per_second": 4.817,
|
| 123 |
+
"eval_steps_per_second": 1.204,
|
| 124 |
+
"step": 68
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.0588235294117645,
|
| 128 |
+
"grad_norm": 0.09218032658100128,
|
| 129 |
+
"learning_rate": 0.0019324722294043557,
|
| 130 |
+
"loss": 1.0691,
|
| 131 |
+
"step": 70
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.2058823529411766,
|
| 135 |
+
"grad_norm": 0.09160086512565613,
|
| 136 |
+
"learning_rate": 0.0019127083684499803,
|
| 137 |
+
"loss": 1.0627,
|
| 138 |
+
"step": 75
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.3529411764705883,
|
| 142 |
+
"grad_norm": 0.08734273165464401,
|
| 143 |
+
"learning_rate": 0.0018905399588691164,
|
| 144 |
+
"loss": 1.0554,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 2.5,
|
| 149 |
+
"grad_norm": 0.16841983795166016,
|
| 150 |
+
"learning_rate": 0.001866025403784439,
|
| 151 |
+
"loss": 1.0504,
|
| 152 |
+
"step": 85
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.6470588235294117,
|
| 156 |
+
"grad_norm": 0.0975656732916832,
|
| 157 |
+
"learning_rate": 0.0018392292872863268,
|
| 158 |
+
"loss": 1.0538,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.7941176470588234,
|
| 163 |
+
"grad_norm": 0.08383582532405853,
|
| 164 |
+
"learning_rate": 0.0018102222042847736,
|
| 165 |
+
"loss": 1.0443,
|
| 166 |
+
"step": 95
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 2.9411764705882355,
|
| 170 |
+
"grad_norm": 0.09132234007120132,
|
| 171 |
+
"learning_rate": 0.0017790805745256705,
|
| 172 |
+
"loss": 1.0447,
|
| 173 |
+
"step": 100
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.0,
|
| 177 |
+
"eval_loss": 1.6731293201446533,
|
| 178 |
+
"eval_runtime": 0.8295,
|
| 179 |
+
"eval_samples_per_second": 4.822,
|
| 180 |
+
"eval_steps_per_second": 1.206,
|
| 181 |
+
"step": 102
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 3.088235294117647,
|
| 185 |
+
"grad_norm": 0.09759514033794403,
|
| 186 |
+
"learning_rate": 0.0017458864412614435,
|
| 187 |
+
"loss": 1.0375,
|
| 188 |
+
"step": 105
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.235294117647059,
|
| 192 |
+
"grad_norm": 0.09400475025177002,
|
| 193 |
+
"learning_rate": 0.0017107272551064472,
|
| 194 |
+
"loss": 1.0277,
|
| 195 |
+
"step": 110
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.3823529411764706,
|
| 199 |
+
"grad_norm": 0.07706195116043091,
|
| 200 |
+
"learning_rate": 0.0016736956436465573,
|
| 201 |
+
"loss": 1.0239,
|
| 202 |
+
"step": 115
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 3.5294117647058822,
|
| 206 |
+
"grad_norm": 0.08647977560758591,
|
| 207 |
+
"learning_rate": 0.0016348891674099228,
|
| 208 |
+
"loss": 1.0229,
|
| 209 |
+
"step": 120
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 3.6764705882352944,
|
| 213 |
+
"grad_norm": 0.08289068937301636,
|
| 214 |
+
"learning_rate": 0.0015944100628417868,
|
| 215 |
+
"loss": 1.0238,
|
| 216 |
+
"step": 125
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 3.8235294117647056,
|
| 220 |
+
"grad_norm": 0.08053147792816162,
|
| 221 |
+
"learning_rate": 0.0015523649729605059,
|
| 222 |
+
"loss": 1.0195,
|
| 223 |
+
"step": 130
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 3.9705882352941178,
|
| 227 |
+
"grad_norm": 0.08788559585809708,
|
| 228 |
+
"learning_rate": 0.001508864666404365,
|
| 229 |
+
"loss": 1.0207,
|
| 230 |
+
"step": 135
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 4.0,
|
| 234 |
+
"eval_loss": 1.6660211086273193,
|
| 235 |
+
"eval_runtime": 0.8294,
|
| 236 |
+
"eval_samples_per_second": 4.823,
|
| 237 |
+
"eval_steps_per_second": 1.206,
|
| 238 |
+
"step": 136
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 4.117647058823529,
|
| 242 |
+
"grad_norm": 0.0886395052075386,
|
| 243 |
+
"learning_rate": 0.0014640237456093634,
|
| 244 |
+
"loss": 1.007,
|
| 245 |
+
"step": 140
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 4.264705882352941,
|
| 249 |
+
"grad_norm": 0.08254272490739822,
|
| 250 |
+
"learning_rate": 0.0014179603448867834,
|
| 251 |
+
"loss": 1.0049,
|
| 252 |
+
"step": 145
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.411764705882353,
|
| 256 |
+
"grad_norm": 0.08430016040802002,
|
| 257 |
+
"learning_rate": 0.0013707958191959608,
|
| 258 |
+
"loss": 1.0103,
|
| 259 |
+
"step": 150
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 4.5588235294117645,
|
| 263 |
+
"grad_norm": 0.0816187709569931,
|
| 264 |
+
"learning_rate": 0.001322654424432195,
|
| 265 |
+
"loss": 1.0011,
|
| 266 |
+
"step": 155
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 4.705882352941177,
|
| 270 |
+
"grad_norm": 0.08980533480644226,
|
| 271 |
+
"learning_rate": 0.0012736629900720832,
|
| 272 |
+
"loss": 1.0107,
|
| 273 |
+
"step": 160
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 4.852941176470588,
|
| 277 |
+
"grad_norm": 0.08757297694683075,
|
| 278 |
+
"learning_rate": 0.0012239505850387032,
|
| 279 |
+
"loss": 1.0035,
|
| 280 |
+
"step": 165
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 5.0,
|
| 284 |
+
"grad_norm": 0.08535553514957428,
|
| 285 |
+
"learning_rate": 0.0011736481776669307,
|
| 286 |
+
"loss": 1.0039,
|
| 287 |
+
"step": 170
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 5.0,
|
| 291 |
+
"eval_loss": 1.668144941329956,
|
| 292 |
+
"eval_runtime": 0.8286,
|
| 293 |
+
"eval_samples_per_second": 4.827,
|
| 294 |
+
"eval_steps_per_second": 1.207,
|
| 295 |
+
"step": 170
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 5.147058823529412,
|
| 299 |
+
"grad_norm": 0.08561732620000839,
|
| 300 |
+
"learning_rate": 0.0011228882906647141,
|
| 301 |
+
"loss": 0.9888,
|
| 302 |
+
"step": 175
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 5.294117647058823,
|
| 306 |
+
"grad_norm": 0.08237646520137787,
|
| 307 |
+
"learning_rate": 0.0010718046519793277,
|
| 308 |
+
"loss": 0.994,
|
| 309 |
+
"step": 180
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.4411764705882355,
|
| 313 |
+
"grad_norm": 0.07582972943782806,
|
| 314 |
+
"learning_rate": 0.0010205318424883906,
|
| 315 |
+
"loss": 0.9953,
|
| 316 |
+
"step": 185
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.588235294117647,
|
| 320 |
+
"grad_norm": 0.07832607626914978,
|
| 321 |
+
"learning_rate": 0.0009692049414438299,
|
| 322 |
+
"loss": 0.991,
|
| 323 |
+
"step": 190
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 5.735294117647059,
|
| 327 |
+
"grad_norm": 0.08117620646953583,
|
| 328 |
+
"learning_rate": 0.0009179591706028624,
|
| 329 |
+
"loss": 0.9931,
|
| 330 |
+
"step": 195
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 5.882352941176471,
|
| 334 |
+
"grad_norm": 0.07949723303318024,
|
| 335 |
+
"learning_rate": 0.0008669295379835467,
|
| 336 |
+
"loss": 0.9957,
|
| 337 |
+
"step": 200
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 6.0,
|
| 341 |
+
"eval_loss": 1.6619951725006104,
|
| 342 |
+
"eval_runtime": 0.8305,
|
| 343 |
+
"eval_samples_per_second": 4.817,
|
| 344 |
+
"eval_steps_per_second": 1.204,
|
| 345 |
+
"step": 204
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.029411764705882,
|
| 349 |
+
"grad_norm": 0.07223918288946152,
|
| 350 |
+
"learning_rate": 0.0008162504821834296,
|
| 351 |
+
"loss": 0.9855,
|
| 352 |
+
"step": 205
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.176470588235294,
|
| 356 |
+
"grad_norm": 0.0734533816576004,
|
| 357 |
+
"learning_rate": 0.0007660555181983517,
|
| 358 |
+
"loss": 0.9822,
|
| 359 |
+
"step": 210
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.323529411764706,
|
| 363 |
+
"grad_norm": 0.07729926705360413,
|
| 364 |
+
"learning_rate": 0.0007164768856744892,
|
| 365 |
+
"loss": 0.9813,
|
| 366 |
+
"step": 215
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.470588235294118,
|
| 370 |
+
"grad_norm": 0.0778665617108345,
|
| 371 |
+
"learning_rate": 0.0006676452005203405,
|
| 372 |
+
"loss": 0.9853,
|
| 373 |
+
"step": 220
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.617647058823529,
|
| 377 |
+
"grad_norm": 0.06967757642269135,
|
| 378 |
+
"learning_rate": 0.0006196891107964744,
|
| 379 |
+
"loss": 0.9837,
|
| 380 |
+
"step": 225
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.764705882352941,
|
| 384 |
+
"grad_norm": 0.07201401889324188,
|
| 385 |
+
"learning_rate": 0.0005727349577896194,
|
| 386 |
+
"loss": 0.9823,
|
| 387 |
+
"step": 230
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.911764705882353,
|
| 391 |
+
"grad_norm": 0.07436826825141907,
|
| 392 |
+
"learning_rate": 0.00052690644316399,
|
| 393 |
+
"loss": 0.9793,
|
| 394 |
+
"step": 235
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 7.0,
|
| 398 |
+
"eval_loss": 1.6655919551849365,
|
| 399 |
+
"eval_runtime": 0.8312,
|
| 400 |
+
"eval_samples_per_second": 4.812,
|
| 401 |
+
"eval_steps_per_second": 1.203,
|
| 402 |
+
"step": 238
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.0588235294117645,
|
| 406 |
+
"grad_norm": 0.07127279043197632,
|
| 407 |
+
"learning_rate": 0.0004823243030667576,
|
| 408 |
+
"loss": 0.98,
|
| 409 |
+
"step": 240
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 7.205882352941177,
|
| 413 |
+
"grad_norm": 0.06584794819355011,
|
| 414 |
+
"learning_rate": 0.0004391059900462304,
|
| 415 |
+
"loss": 0.9738,
|
| 416 |
+
"step": 245
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.352941176470588,
|
| 420 |
+
"grad_norm": 0.06734811514616013,
|
| 421 |
+
"learning_rate": 0.0003973653636207437,
|
| 422 |
+
"loss": 0.9705,
|
| 423 |
+
"step": 250
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.5,
|
| 427 |
+
"grad_norm": 0.06853578239679337,
|
| 428 |
+
"learning_rate": 0.0003572123903134606,
|
| 429 |
+
"loss": 0.9778,
|
| 430 |
+
"step": 255
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.647058823529412,
|
| 434 |
+
"grad_norm": 0.0678030252456665,
|
| 435 |
+
"learning_rate": 0.0003187528539433457,
|
| 436 |
+
"loss": 0.9722,
|
| 437 |
+
"step": 260
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.794117647058823,
|
| 441 |
+
"grad_norm": 0.06483301520347595,
|
| 442 |
+
"learning_rate": 0.0002820880769355582,
|
| 443 |
+
"loss": 0.9782,
|
| 444 |
+
"step": 265
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.9411764705882355,
|
| 448 |
+
"grad_norm": 0.06444835662841797,
|
| 449 |
+
"learning_rate": 0.00024731465338547555,
|
| 450 |
+
"loss": 0.9761,
|
| 451 |
+
"step": 270
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 8.0,
|
| 455 |
+
"eval_loss": 1.670668125152588,
|
| 456 |
+
"eval_runtime": 0.8308,
|
| 457 |
+
"eval_samples_per_second": 4.815,
|
| 458 |
+
"eval_steps_per_second": 1.204,
|
| 459 |
+
"step": 272
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 8.088235294117647,
|
| 463 |
+
"grad_norm": 0.06315235048532486,
|
| 464 |
+
"learning_rate": 0.00021452419457960138,
|
| 465 |
+
"loss": 0.9718,
|
| 466 |
+
"step": 275
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.235294117647058,
|
| 470 |
+
"grad_norm": 0.06345119327306747,
|
| 471 |
+
"learning_rate": 0.0001838030876437784,
|
| 472 |
+
"loss": 0.977,
|
| 473 |
+
"step": 280
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 8.382352941176471,
|
| 477 |
+
"grad_norm": 0.0637635886669159,
|
| 478 |
+
"learning_rate": 0.00015523226795456348,
|
| 479 |
+
"loss": 0.9741,
|
| 480 |
+
"step": 285
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 8.529411764705882,
|
| 484 |
+
"grad_norm": 0.06295765936374664,
|
| 485 |
+
"learning_rate": 0.00012888700591334225,
|
| 486 |
+
"loss": 0.9698,
|
| 487 |
+
"step": 290
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 8.676470588235293,
|
| 491 |
+
"grad_norm": 0.06005469709634781,
|
| 492 |
+
"learning_rate": 0.00010483670864493777,
|
| 493 |
+
"loss": 0.967,
|
| 494 |
+
"step": 295
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 8.823529411764707,
|
| 498 |
+
"grad_norm": 0.060096003115177155,
|
| 499 |
+
"learning_rate": 8.31447371431372e-05,
|
| 500 |
+
"loss": 0.9651,
|
| 501 |
+
"step": 300
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 8.970588235294118,
|
| 505 |
+
"grad_norm": 0.05939820781350136,
|
| 506 |
+
"learning_rate": 6.386823934487617e-05,
|
| 507 |
+
"loss": 0.9678,
|
| 508 |
+
"step": 305
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 9.0,
|
| 512 |
+
"eval_loss": 1.6740888357162476,
|
| 513 |
+
"eval_runtime": 0.8293,
|
| 514 |
+
"eval_samples_per_second": 4.824,
|
| 515 |
+
"eval_steps_per_second": 1.206,
|
| 516 |
+
"step": 306
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 9.117647058823529,
|
| 520 |
+
"grad_norm": 0.058787424117326736,
|
| 521 |
+
"learning_rate": 4.705799957284351e-05,
|
| 522 |
+
"loss": 0.9715,
|
| 523 |
+
"step": 310
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 9.264705882352942,
|
| 527 |
+
"grad_norm": 0.057846549898386,
|
| 528 |
+
"learning_rate": 3.275830474315855e-05,
|
| 529 |
+
"loss": 0.9632,
|
| 530 |
+
"step": 315
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.411764705882353,
|
| 534 |
+
"grad_norm": 0.05796463415026665,
|
| 535 |
+
"learning_rate": 2.1006827690595476e-05,
|
| 536 |
+
"loss": 0.9622,
|
| 537 |
+
"step": 320
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 9.558823529411764,
|
| 541 |
+
"grad_norm": 0.057375218719244,
|
| 542 |
+
"learning_rate": 1.1834527918740623e-05,
|
| 543 |
+
"loss": 0.9712,
|
| 544 |
+
"step": 325
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 9.705882352941176,
|
| 548 |
+
"grad_norm": 0.05756480246782303,
|
| 549 |
+
"learning_rate": 5.265570036553813e-06,
|
| 550 |
+
"loss": 0.9672,
|
| 551 |
+
"step": 330
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 9.852941176470589,
|
| 555 |
+
"grad_norm": 0.05712839215993881,
|
| 556 |
+
"learning_rate": 1.3172600962190196e-06,
|
| 557 |
+
"loss": 0.9695,
|
| 558 |
+
"step": 335
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 10.0,
|
| 562 |
+
"grad_norm": 0.057304926216602325,
|
| 563 |
+
"learning_rate": 0.0,
|
| 564 |
+
"loss": 0.9709,
|
| 565 |
+
"step": 340
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 10.0,
|
| 569 |
+
"eval_loss": 1.6745353937149048,
|
| 570 |
+
"eval_runtime": 0.8292,
|
| 571 |
+
"eval_samples_per_second": 4.824,
|
| 572 |
+
"eval_steps_per_second": 1.206,
|
| 573 |
+
"step": 340
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 10.0,
|
| 577 |
+
"step": 340,
|
| 578 |
+
"total_flos": 1.0444708917333197e+18,
|
| 579 |
+
"train_loss": 1.0473914388348073,
|
| 580 |
+
"train_runtime": 1591.5084,
|
| 581 |
+
"train_samples_per_second": 108.407,
|
| 582 |
+
"train_steps_per_second": 0.214
|
| 583 |
+
}
|
| 584 |
+
],
|
| 585 |
+
"logging_steps": 5,
|
| 586 |
+
"max_steps": 340,
|
| 587 |
+
"num_input_tokens_seen": 0,
|
| 588 |
+
"num_train_epochs": 10,
|
| 589 |
+
"save_steps": 100,
|
| 590 |
+
"stateful_callbacks": {
|
| 591 |
+
"TrainerControl": {
|
| 592 |
+
"args": {
|
| 593 |
+
"should_epoch_stop": false,
|
| 594 |
+
"should_evaluate": false,
|
| 595 |
+
"should_log": false,
|
| 596 |
+
"should_save": true,
|
| 597 |
+
"should_training_stop": true
|
| 598 |
+
},
|
| 599 |
+
"attributes": {}
|
| 600 |
+
}
|
| 601 |
+
},
|
| 602 |
+
"total_flos": 1.0444708917333197e+18,
|
| 603 |
+
"train_batch_size": 32,
|
| 604 |
+
"trial_name": null,
|
| 605 |
+
"trial_params": null
|
| 606 |
+
}
|