Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
license: llama3.2
|
| 4 |
+
base_model: meta-llama/Llama-3.2-3B
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- sft
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- generator
|
| 11 |
+
model-index:
|
| 12 |
+
- name: llama3-3b-classification-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 |
+
# llama3-3b-classification-gpt4o-100k
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-3B](https://huggingface.co/meta-llama/Llama-3.2-3B) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 2.2580
|
| 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: 16
|
| 44 |
+
- eval_batch_size: 16
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 8
|
| 48 |
+
- gradient_accumulation_steps: 2
|
| 49 |
+
- total_train_batch_size: 256
|
| 50 |
+
- total_eval_batch_size: 128
|
| 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 |
+
| 2.1922 | 1.0 | 30 | 2.2939 |
|
| 61 |
+
| 1.9883 | 2.0 | 60 | 2.2309 |
|
| 62 |
+
| 1.9261 | 3.0 | 90 | 2.2480 |
|
| 63 |
+
| 1.877 | 4.0 | 120 | 2.2464 |
|
| 64 |
+
| 1.8508 | 5.0 | 150 | 2.2514 |
|
| 65 |
+
| 1.8357 | 6.0 | 180 | 2.2442 |
|
| 66 |
+
| 1.8225 | 7.0 | 210 | 2.2535 |
|
| 67 |
+
| 1.8153 | 8.0 | 240 | 2.2560 |
|
| 68 |
+
| 1.8065 | 9.0 | 270 | 2.2574 |
|
| 69 |
+
| 1.8084 | 10.0 | 300 | 2.2580 |
|
| 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
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1612749744
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e9fe231b4e6cb22f135f37cf2ba422fdca942aa24a0f807f335dd194d6c8922
|
| 3 |
size 1612749744
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 1.3343811847501906e+18,
|
| 4 |
+
"train_loss": 1.9245261510213216,
|
| 5 |
+
"train_runtime": 1578.2466,
|
| 6 |
+
"train_samples": 92634,
|
| 7 |
+
"train_samples_per_second": 47.958,
|
| 8 |
+
"train_steps_per_second": 0.19
|
| 9 |
+
}
|
runs/Apr01_00-42-00_green-face-echoes-fin-01/events.out.tfevents.1743468298.green-face-echoes-fin-01.8410.0
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:db9596aa9ef447df0d8b1f695cd826840254e5eea46b5b5759dfd386279e5ec8
|
| 3 |
+
size 22101
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 1.3343811847501906e+18,
|
| 4 |
+
"train_loss": 1.9245261510213216,
|
| 5 |
+
"train_runtime": 1578.2466,
|
| 6 |
+
"train_samples": 92634,
|
| 7 |
+
"train_samples_per_second": 47.958,
|
| 8 |
+
"train_steps_per_second": 0.19
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,550 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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": 300,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03333333333333333,
|
| 14 |
+
"grad_norm": 0.8721705675125122,
|
| 15 |
+
"learning_rate": 6.666666666666667e-06,
|
| 16 |
+
"loss": 2.5652,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16666666666666666,
|
| 21 |
+
"grad_norm": 0.7771564722061157,
|
| 22 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 23 |
+
"loss": 2.5757,
|
| 24 |
+
"step": 5
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3333333333333333,
|
| 28 |
+
"grad_norm": 0.4726114273071289,
|
| 29 |
+
"learning_rate": 6.666666666666667e-05,
|
| 30 |
+
"loss": 2.5393,
|
| 31 |
+
"step": 10
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5,
|
| 35 |
+
"grad_norm": 0.41005024313926697,
|
| 36 |
+
"learning_rate": 0.0001,
|
| 37 |
+
"loss": 2.4619,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6666666666666666,
|
| 42 |
+
"grad_norm": 0.359015554189682,
|
| 43 |
+
"learning_rate": 0.00013333333333333334,
|
| 44 |
+
"loss": 2.3546,
|
| 45 |
+
"step": 20
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8333333333333334,
|
| 49 |
+
"grad_norm": 0.32825127243995667,
|
| 50 |
+
"learning_rate": 0.0001666666666666667,
|
| 51 |
+
"loss": 2.2723,
|
| 52 |
+
"step": 25
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"grad_norm": 0.2509020268917084,
|
| 57 |
+
"learning_rate": 0.0002,
|
| 58 |
+
"loss": 2.1922,
|
| 59 |
+
"step": 30
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_loss": 2.293870449066162,
|
| 64 |
+
"eval_runtime": 0.868,
|
| 65 |
+
"eval_samples_per_second": 2.304,
|
| 66 |
+
"eval_steps_per_second": 1.152,
|
| 67 |
+
"step": 30
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.1666666666666667,
|
| 71 |
+
"grad_norm": 0.17683961987495422,
|
| 72 |
+
"learning_rate": 0.00019983081582712685,
|
| 73 |
+
"loss": 2.1372,
|
| 74 |
+
"step": 35
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.3333333333333333,
|
| 78 |
+
"grad_norm": 0.13620507717132568,
|
| 79 |
+
"learning_rate": 0.00019932383577419432,
|
| 80 |
+
"loss": 2.092,
|
| 81 |
+
"step": 40
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.5,
|
| 85 |
+
"grad_norm": 0.1294490098953247,
|
| 86 |
+
"learning_rate": 0.00019848077530122083,
|
| 87 |
+
"loss": 2.0612,
|
| 88 |
+
"step": 45
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.6666666666666665,
|
| 92 |
+
"grad_norm": 0.12196648120880127,
|
| 93 |
+
"learning_rate": 0.00019730448705798239,
|
| 94 |
+
"loss": 2.029,
|
| 95 |
+
"step": 50
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.8333333333333335,
|
| 99 |
+
"grad_norm": 0.12746499478816986,
|
| 100 |
+
"learning_rate": 0.0001957989512315489,
|
| 101 |
+
"loss": 2.0055,
|
| 102 |
+
"step": 55
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.0,
|
| 106 |
+
"grad_norm": 0.11664091795682907,
|
| 107 |
+
"learning_rate": 0.00019396926207859084,
|
| 108 |
+
"loss": 1.9883,
|
| 109 |
+
"step": 60
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.0,
|
| 113 |
+
"eval_loss": 2.2309093475341797,
|
| 114 |
+
"eval_runtime": 0.8673,
|
| 115 |
+
"eval_samples_per_second": 2.306,
|
| 116 |
+
"eval_steps_per_second": 1.153,
|
| 117 |
+
"step": 60
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 2.1666666666666665,
|
| 121 |
+
"grad_norm": 0.12395822256803513,
|
| 122 |
+
"learning_rate": 0.00019182161068802741,
|
| 123 |
+
"loss": 1.9631,
|
| 124 |
+
"step": 65
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.3333333333333335,
|
| 128 |
+
"grad_norm": 0.14801518619060516,
|
| 129 |
+
"learning_rate": 0.00018936326403234125,
|
| 130 |
+
"loss": 1.9482,
|
| 131 |
+
"step": 70
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.5,
|
| 135 |
+
"grad_norm": 0.12537938356399536,
|
| 136 |
+
"learning_rate": 0.00018660254037844388,
|
| 137 |
+
"loss": 1.9543,
|
| 138 |
+
"step": 75
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.6666666666666665,
|
| 142 |
+
"grad_norm": 0.13340197503566742,
|
| 143 |
+
"learning_rate": 0.00018354878114129367,
|
| 144 |
+
"loss": 1.9364,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 2.8333333333333335,
|
| 149 |
+
"grad_norm": 0.16110606491565704,
|
| 150 |
+
"learning_rate": 0.0001802123192755044,
|
| 151 |
+
"loss": 1.9286,
|
| 152 |
+
"step": 85
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.0,
|
| 156 |
+
"grad_norm": 0.14026618003845215,
|
| 157 |
+
"learning_rate": 0.0001766044443118978,
|
| 158 |
+
"loss": 1.9261,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 3.0,
|
| 163 |
+
"eval_loss": 2.248041868209839,
|
| 164 |
+
"eval_runtime": 0.8685,
|
| 165 |
+
"eval_samples_per_second": 2.303,
|
| 166 |
+
"eval_steps_per_second": 1.151,
|
| 167 |
+
"step": 90
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 3.1666666666666665,
|
| 171 |
+
"grad_norm": 0.13984443247318268,
|
| 172 |
+
"learning_rate": 0.00017273736415730488,
|
| 173 |
+
"loss": 1.9053,
|
| 174 |
+
"step": 95
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 3.3333333333333335,
|
| 178 |
+
"grad_norm": 0.17301388084888458,
|
| 179 |
+
"learning_rate": 0.0001686241637868734,
|
| 180 |
+
"loss": 1.9071,
|
| 181 |
+
"step": 100
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 3.5,
|
| 185 |
+
"grad_norm": 0.15998174250125885,
|
| 186 |
+
"learning_rate": 0.00016427876096865394,
|
| 187 |
+
"loss": 1.896,
|
| 188 |
+
"step": 105
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.6666666666666665,
|
| 192 |
+
"grad_norm": 0.17066697776317596,
|
| 193 |
+
"learning_rate": 0.00015971585917027862,
|
| 194 |
+
"loss": 1.891,
|
| 195 |
+
"step": 110
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.8333333333333335,
|
| 199 |
+
"grad_norm": 0.18711982667446136,
|
| 200 |
+
"learning_rate": 0.0001549508978070806,
|
| 201 |
+
"loss": 1.8784,
|
| 202 |
+
"step": 115
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 4.0,
|
| 206 |
+
"grad_norm": 0.18189194798469543,
|
| 207 |
+
"learning_rate": 0.00015000000000000001,
|
| 208 |
+
"loss": 1.877,
|
| 209 |
+
"step": 120
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 4.0,
|
| 213 |
+
"eval_loss": 2.246403217315674,
|
| 214 |
+
"eval_runtime": 0.8876,
|
| 215 |
+
"eval_samples_per_second": 2.253,
|
| 216 |
+
"eval_steps_per_second": 1.127,
|
| 217 |
+
"step": 120
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 4.166666666666667,
|
| 221 |
+
"grad_norm": 0.19101205468177795,
|
| 222 |
+
"learning_rate": 0.00014487991802004623,
|
| 223 |
+
"loss": 1.8693,
|
| 224 |
+
"step": 125
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 4.333333333333333,
|
| 228 |
+
"grad_norm": 0.18224620819091797,
|
| 229 |
+
"learning_rate": 0.0001396079766039157,
|
| 230 |
+
"loss": 1.8717,
|
| 231 |
+
"step": 130
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 4.5,
|
| 235 |
+
"grad_norm": 0.21038778126239777,
|
| 236 |
+
"learning_rate": 0.00013420201433256689,
|
| 237 |
+
"loss": 1.865,
|
| 238 |
+
"step": 135
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 4.666666666666667,
|
| 242 |
+
"grad_norm": 0.20972222089767456,
|
| 243 |
+
"learning_rate": 0.00012868032327110904,
|
| 244 |
+
"loss": 1.8534,
|
| 245 |
+
"step": 140
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 4.833333333333333,
|
| 249 |
+
"grad_norm": 0.20562510192394257,
|
| 250 |
+
"learning_rate": 0.00012306158707424403,
|
| 251 |
+
"loss": 1.8507,
|
| 252 |
+
"step": 145
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 5.0,
|
| 256 |
+
"grad_norm": 0.18542522192001343,
|
| 257 |
+
"learning_rate": 0.00011736481776669306,
|
| 258 |
+
"loss": 1.8508,
|
| 259 |
+
"step": 150
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 5.0,
|
| 263 |
+
"eval_loss": 2.251426935195923,
|
| 264 |
+
"eval_runtime": 0.8697,
|
| 265 |
+
"eval_samples_per_second": 2.3,
|
| 266 |
+
"eval_steps_per_second": 1.15,
|
| 267 |
+
"step": 150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 5.166666666666667,
|
| 271 |
+
"grad_norm": 0.19300299882888794,
|
| 272 |
+
"learning_rate": 0.00011160929141252303,
|
| 273 |
+
"loss": 1.8441,
|
| 274 |
+
"step": 155
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 5.333333333333333,
|
| 278 |
+
"grad_norm": 0.17082081735134125,
|
| 279 |
+
"learning_rate": 0.00010581448289104758,
|
| 280 |
+
"loss": 1.8406,
|
| 281 |
+
"step": 160
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 5.5,
|
| 285 |
+
"grad_norm": 0.16670560836791992,
|
| 286 |
+
"learning_rate": 0.0001,
|
| 287 |
+
"loss": 1.8314,
|
| 288 |
+
"step": 165
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 5.666666666666667,
|
| 292 |
+
"grad_norm": 0.2005225419998169,
|
| 293 |
+
"learning_rate": 9.418551710895243e-05,
|
| 294 |
+
"loss": 1.84,
|
| 295 |
+
"step": 170
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 5.833333333333333,
|
| 299 |
+
"grad_norm": 0.17486542463302612,
|
| 300 |
+
"learning_rate": 8.839070858747697e-05,
|
| 301 |
+
"loss": 1.8356,
|
| 302 |
+
"step": 175
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 6.0,
|
| 306 |
+
"grad_norm": 0.17715969681739807,
|
| 307 |
+
"learning_rate": 8.263518223330697e-05,
|
| 308 |
+
"loss": 1.8357,
|
| 309 |
+
"step": 180
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 6.0,
|
| 313 |
+
"eval_loss": 2.244203805923462,
|
| 314 |
+
"eval_runtime": 0.8698,
|
| 315 |
+
"eval_samples_per_second": 2.299,
|
| 316 |
+
"eval_steps_per_second": 1.15,
|
| 317 |
+
"step": 180
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.166666666666667,
|
| 321 |
+
"grad_norm": 0.1805175244808197,
|
| 322 |
+
"learning_rate": 7.693841292575598e-05,
|
| 323 |
+
"loss": 1.8293,
|
| 324 |
+
"step": 185
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.333333333333333,
|
| 328 |
+
"grad_norm": 0.16339732706546783,
|
| 329 |
+
"learning_rate": 7.131967672889101e-05,
|
| 330 |
+
"loss": 1.8289,
|
| 331 |
+
"step": 190
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.5,
|
| 335 |
+
"grad_norm": 0.17380179464817047,
|
| 336 |
+
"learning_rate": 6.579798566743314e-05,
|
| 337 |
+
"loss": 1.8242,
|
| 338 |
+
"step": 195
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.666666666666667,
|
| 342 |
+
"grad_norm": 0.18039536476135254,
|
| 343 |
+
"learning_rate": 6.039202339608432e-05,
|
| 344 |
+
"loss": 1.8221,
|
| 345 |
+
"step": 200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.833333333333333,
|
| 349 |
+
"grad_norm": 0.1678876280784607,
|
| 350 |
+
"learning_rate": 5.5120081979953785e-05,
|
| 351 |
+
"loss": 1.8182,
|
| 352 |
+
"step": 205
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.0,
|
| 356 |
+
"grad_norm": 0.18717928230762482,
|
| 357 |
+
"learning_rate": 5.000000000000002e-05,
|
| 358 |
+
"loss": 1.8225,
|
| 359 |
+
"step": 210
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.0,
|
| 363 |
+
"eval_loss": 2.2534749507904053,
|
| 364 |
+
"eval_runtime": 0.8691,
|
| 365 |
+
"eval_samples_per_second": 2.301,
|
| 366 |
+
"eval_steps_per_second": 1.151,
|
| 367 |
+
"step": 210
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 7.166666666666667,
|
| 371 |
+
"grad_norm": 0.1713215708732605,
|
| 372 |
+
"learning_rate": 4.50491021929194e-05,
|
| 373 |
+
"loss": 1.8162,
|
| 374 |
+
"step": 215
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 7.333333333333333,
|
| 378 |
+
"grad_norm": 0.14658918976783752,
|
| 379 |
+
"learning_rate": 4.028414082972141e-05,
|
| 380 |
+
"loss": 1.8146,
|
| 381 |
+
"step": 220
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 7.5,
|
| 385 |
+
"grad_norm": 0.16408003866672516,
|
| 386 |
+
"learning_rate": 3.5721239031346066e-05,
|
| 387 |
+
"loss": 1.8179,
|
| 388 |
+
"step": 225
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 7.666666666666667,
|
| 392 |
+
"grad_norm": 0.15933865308761597,
|
| 393 |
+
"learning_rate": 3.137583621312665e-05,
|
| 394 |
+
"loss": 1.8101,
|
| 395 |
+
"step": 230
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 7.833333333333333,
|
| 399 |
+
"grad_norm": 0.14181320369243622,
|
| 400 |
+
"learning_rate": 2.7262635842695127e-05,
|
| 401 |
+
"loss": 1.8199,
|
| 402 |
+
"step": 235
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 8.0,
|
| 406 |
+
"grad_norm": 0.16157026588916779,
|
| 407 |
+
"learning_rate": 2.339555568810221e-05,
|
| 408 |
+
"loss": 1.8153,
|
| 409 |
+
"step": 240
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 8.0,
|
| 413 |
+
"eval_loss": 2.2560064792633057,
|
| 414 |
+
"eval_runtime": 0.8693,
|
| 415 |
+
"eval_samples_per_second": 2.301,
|
| 416 |
+
"eval_steps_per_second": 1.15,
|
| 417 |
+
"step": 240
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 8.166666666666666,
|
| 421 |
+
"grad_norm": 0.14123442769050598,
|
| 422 |
+
"learning_rate": 1.9787680724495617e-05,
|
| 423 |
+
"loss": 1.8138,
|
| 424 |
+
"step": 245
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 8.333333333333334,
|
| 428 |
+
"grad_norm": 0.14202268421649933,
|
| 429 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 430 |
+
"loss": 1.8143,
|
| 431 |
+
"step": 250
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 8.5,
|
| 435 |
+
"grad_norm": 0.15732887387275696,
|
| 436 |
+
"learning_rate": 1.339745962155613e-05,
|
| 437 |
+
"loss": 1.811,
|
| 438 |
+
"step": 255
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.666666666666666,
|
| 442 |
+
"grad_norm": 0.13898594677448273,
|
| 443 |
+
"learning_rate": 1.0636735967658784e-05,
|
| 444 |
+
"loss": 1.8033,
|
| 445 |
+
"step": 260
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 8.833333333333334,
|
| 449 |
+
"grad_norm": 0.1528938263654709,
|
| 450 |
+
"learning_rate": 8.178389311972612e-06,
|
| 451 |
+
"loss": 1.8194,
|
| 452 |
+
"step": 265
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 9.0,
|
| 456 |
+
"grad_norm": 0.14544987678527832,
|
| 457 |
+
"learning_rate": 6.030737921409169e-06,
|
| 458 |
+
"loss": 1.8065,
|
| 459 |
+
"step": 270
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 9.0,
|
| 463 |
+
"eval_loss": 2.257399320602417,
|
| 464 |
+
"eval_runtime": 0.8685,
|
| 465 |
+
"eval_samples_per_second": 2.303,
|
| 466 |
+
"eval_steps_per_second": 1.151,
|
| 467 |
+
"step": 270
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 9.166666666666666,
|
| 471 |
+
"grad_norm": 0.13697919249534607,
|
| 472 |
+
"learning_rate": 4.20104876845111e-06,
|
| 473 |
+
"loss": 1.8037,
|
| 474 |
+
"step": 275
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 9.333333333333334,
|
| 478 |
+
"grad_norm": 0.134043887257576,
|
| 479 |
+
"learning_rate": 2.6955129420176196e-06,
|
| 480 |
+
"loss": 1.8095,
|
| 481 |
+
"step": 280
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 9.5,
|
| 485 |
+
"grad_norm": 0.1406584531068802,
|
| 486 |
+
"learning_rate": 1.5192246987791981e-06,
|
| 487 |
+
"loss": 1.8133,
|
| 488 |
+
"step": 285
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 9.666666666666666,
|
| 492 |
+
"grad_norm": 0.1350001096725464,
|
| 493 |
+
"learning_rate": 6.761642258056978e-07,
|
| 494 |
+
"loss": 1.8064,
|
| 495 |
+
"step": 290
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 9.833333333333334,
|
| 499 |
+
"grad_norm": 0.13039974868297577,
|
| 500 |
+
"learning_rate": 1.6918417287318245e-07,
|
| 501 |
+
"loss": 1.8157,
|
| 502 |
+
"step": 295
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 10.0,
|
| 506 |
+
"grad_norm": 0.12890292704105377,
|
| 507 |
+
"learning_rate": 0.0,
|
| 508 |
+
"loss": 1.8084,
|
| 509 |
+
"step": 300
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 10.0,
|
| 513 |
+
"eval_loss": 2.2579641342163086,
|
| 514 |
+
"eval_runtime": 0.8884,
|
| 515 |
+
"eval_samples_per_second": 2.251,
|
| 516 |
+
"eval_steps_per_second": 1.126,
|
| 517 |
+
"step": 300
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 10.0,
|
| 521 |
+
"step": 300,
|
| 522 |
+
"total_flos": 1.3343811847501906e+18,
|
| 523 |
+
"train_loss": 1.9245261510213216,
|
| 524 |
+
"train_runtime": 1578.2466,
|
| 525 |
+
"train_samples_per_second": 47.958,
|
| 526 |
+
"train_steps_per_second": 0.19
|
| 527 |
+
}
|
| 528 |
+
],
|
| 529 |
+
"logging_steps": 5,
|
| 530 |
+
"max_steps": 300,
|
| 531 |
+
"num_input_tokens_seen": 0,
|
| 532 |
+
"num_train_epochs": 10,
|
| 533 |
+
"save_steps": 100,
|
| 534 |
+
"stateful_callbacks": {
|
| 535 |
+
"TrainerControl": {
|
| 536 |
+
"args": {
|
| 537 |
+
"should_epoch_stop": false,
|
| 538 |
+
"should_evaluate": false,
|
| 539 |
+
"should_log": false,
|
| 540 |
+
"should_save": true,
|
| 541 |
+
"should_training_stop": true
|
| 542 |
+
},
|
| 543 |
+
"attributes": {}
|
| 544 |
+
}
|
| 545 |
+
},
|
| 546 |
+
"total_flos": 1.3343811847501906e+18,
|
| 547 |
+
"train_batch_size": 16,
|
| 548 |
+
"trial_name": null,
|
| 549 |
+
"trial_params": null
|
| 550 |
+
}
|