Commit
•
c8bc8c4
1
Parent(s):
abd54a7
Model save
Browse files- README.md +98 -0
- adapter_model.safetensors +1 -1
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +1338 -0
README.md
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
|
| 3 |
+
library_name: peft
|
| 4 |
+
license: llama3.1
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- sft
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
model-index:
|
| 10 |
+
- name: Llama-31-8B_task-3_180-samples_config-4
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# Llama-31-8B_task-3_180-samples_config-4
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the None dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.5094
|
| 22 |
+
|
| 23 |
+
## Model description
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Intended uses & limitations
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training and evaluation data
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Training procedure
|
| 36 |
+
|
| 37 |
+
### Training hyperparameters
|
| 38 |
+
|
| 39 |
+
The following hyperparameters were used during training:
|
| 40 |
+
- learning_rate: 1e-05
|
| 41 |
+
- train_batch_size: 1
|
| 42 |
+
- eval_batch_size: 1
|
| 43 |
+
- seed: 42
|
| 44 |
+
- distributed_type: multi-GPU
|
| 45 |
+
- gradient_accumulation_steps: 16
|
| 46 |
+
- total_train_batch_size: 16
|
| 47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
+
- lr_scheduler_type: cosine
|
| 49 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 50 |
+
- num_epochs: 150
|
| 51 |
+
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 55 |
+
|:-------------:|:-------:|:----:|:---------------:|
|
| 56 |
+
| 2.5495 | 0.9412 | 8 | 2.5077 |
|
| 57 |
+
| 2.3733 | 2.0 | 17 | 2.4775 |
|
| 58 |
+
| 2.46 | 2.9412 | 25 | 2.4218 |
|
| 59 |
+
| 2.4585 | 4.0 | 34 | 2.3278 |
|
| 60 |
+
| 2.2624 | 4.9412 | 42 | 2.1919 |
|
| 61 |
+
| 2.0553 | 6.0 | 51 | 1.9704 |
|
| 62 |
+
| 1.7403 | 6.9412 | 59 | 1.7066 |
|
| 63 |
+
| 1.3756 | 8.0 | 68 | 1.3617 |
|
| 64 |
+
| 1.11 | 8.9412 | 76 | 1.0613 |
|
| 65 |
+
| 0.7161 | 10.0 | 85 | 0.7772 |
|
| 66 |
+
| 0.7609 | 10.9412 | 93 | 0.6787 |
|
| 67 |
+
| 0.4358 | 12.0 | 102 | 0.6182 |
|
| 68 |
+
| 0.4774 | 12.9412 | 110 | 0.5912 |
|
| 69 |
+
| 0.5569 | 14.0 | 119 | 0.5746 |
|
| 70 |
+
| 0.427 | 14.9412 | 127 | 0.5487 |
|
| 71 |
+
| 0.4672 | 16.0 | 136 | 0.5339 |
|
| 72 |
+
| 0.3495 | 16.9412 | 144 | 0.5525 |
|
| 73 |
+
| 0.4731 | 18.0 | 153 | 0.5323 |
|
| 74 |
+
| 0.3913 | 18.9412 | 161 | 0.5243 |
|
| 75 |
+
| 0.5624 | 20.0 | 170 | 0.5253 |
|
| 76 |
+
| 0.4684 | 20.9412 | 178 | 0.5222 |
|
| 77 |
+
| 0.3029 | 22.0 | 187 | 0.5100 |
|
| 78 |
+
| 0.3522 | 22.9412 | 195 | 0.5085 |
|
| 79 |
+
| 0.3855 | 24.0 | 204 | 0.4971 |
|
| 80 |
+
| 0.317 | 24.9412 | 212 | 0.5049 |
|
| 81 |
+
| 0.338 | 26.0 | 221 | 0.5016 |
|
| 82 |
+
| 0.391 | 26.9412 | 229 | 0.4942 |
|
| 83 |
+
| 0.3964 | 28.0 | 238 | 0.5010 |
|
| 84 |
+
| 0.2951 | 28.9412 | 246 | 0.5098 |
|
| 85 |
+
| 0.4021 | 30.0 | 255 | 0.5068 |
|
| 86 |
+
| 0.4021 | 30.9412 | 263 | 0.5070 |
|
| 87 |
+
| 0.3456 | 32.0 | 272 | 0.5025 |
|
| 88 |
+
| 0.4431 | 32.9412 | 280 | 0.5050 |
|
| 89 |
+
| 0.4131 | 34.0 | 289 | 0.5094 |
|
| 90 |
+
|
| 91 |
+
|
| 92 |
+
### Framework versions
|
| 93 |
+
|
| 94 |
+
- PEFT 0.12.0
|
| 95 |
+
- Transformers 4.44.0
|
| 96 |
+
- Pytorch 2.1.2+cu121
|
| 97 |
+
- Datasets 2.20.0
|
| 98 |
+
- Tokenizers 0.19.1
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167832240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45be132bd6ab4d2647744dcb7d62c0ad21afc4292b69233df0d5926efa9e3610
|
| 3 |
size 167832240
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 34.0,
|
| 3 |
+
"total_flos": 3.6265283397720474e+17,
|
| 4 |
+
"train_loss": 0.8757064017663777,
|
| 5 |
+
"train_runtime": 13296.8495,
|
| 6 |
+
"train_samples": 136,
|
| 7 |
+
"train_samples_per_second": 1.534,
|
| 8 |
+
"train_steps_per_second": 0.09
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 34.0,
|
| 3 |
+
"total_flos": 3.6265283397720474e+17,
|
| 4 |
+
"train_loss": 0.8757064017663777,
|
| 5 |
+
"train_runtime": 13296.8495,
|
| 6 |
+
"train_samples": 136,
|
| 7 |
+
"train_samples_per_second": 1.534,
|
| 8 |
+
"train_steps_per_second": 0.09
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1338 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.49422508478164673,
|
| 3 |
+
"best_model_checkpoint": "data/Llama-31-8B_task-3_180-samples_config-4/checkpoint-229",
|
| 4 |
+
"epoch": 34.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 289,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11764705882352941,
|
| 13 |
+
"grad_norm": 2.2058663368225098,
|
| 14 |
+
"learning_rate": 8.333333333333334e-08,
|
| 15 |
+
"loss": 2.3266,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.23529411764705882,
|
| 20 |
+
"grad_norm": 3.2251737117767334,
|
| 21 |
+
"learning_rate": 1.6666666666666668e-07,
|
| 22 |
+
"loss": 2.8101,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.47058823529411764,
|
| 27 |
+
"grad_norm": 2.427489757537842,
|
| 28 |
+
"learning_rate": 3.3333333333333335e-07,
|
| 29 |
+
"loss": 2.4918,
|
| 30 |
+
"step": 4
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.7058823529411765,
|
| 34 |
+
"grad_norm": 2.110276460647583,
|
| 35 |
+
"learning_rate": 5.000000000000001e-07,
|
| 36 |
+
"loss": 2.416,
|
| 37 |
+
"step": 6
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.9411764705882353,
|
| 41 |
+
"grad_norm": 2.285691499710083,
|
| 42 |
+
"learning_rate": 6.666666666666667e-07,
|
| 43 |
+
"loss": 2.5495,
|
| 44 |
+
"step": 8
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.9411764705882353,
|
| 48 |
+
"eval_loss": 2.5076982975006104,
|
| 49 |
+
"eval_runtime": 31.8608,
|
| 50 |
+
"eval_samples_per_second": 1.13,
|
| 51 |
+
"eval_steps_per_second": 1.13,
|
| 52 |
+
"step": 8
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.1764705882352942,
|
| 56 |
+
"grad_norm": 2.757939577102661,
|
| 57 |
+
"learning_rate": 8.333333333333333e-07,
|
| 58 |
+
"loss": 2.4129,
|
| 59 |
+
"step": 10
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.4117647058823528,
|
| 63 |
+
"grad_norm": 3.1854565143585205,
|
| 64 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 65 |
+
"loss": 2.6068,
|
| 66 |
+
"step": 12
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.6470588235294117,
|
| 70 |
+
"grad_norm": 2.928661346435547,
|
| 71 |
+
"learning_rate": 1.1666666666666668e-06,
|
| 72 |
+
"loss": 2.66,
|
| 73 |
+
"step": 14
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.8823529411764706,
|
| 77 |
+
"grad_norm": 2.948481321334839,
|
| 78 |
+
"learning_rate": 1.3333333333333334e-06,
|
| 79 |
+
"loss": 2.3733,
|
| 80 |
+
"step": 16
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.0,
|
| 84 |
+
"eval_loss": 2.4775068759918213,
|
| 85 |
+
"eval_runtime": 31.8538,
|
| 86 |
+
"eval_samples_per_second": 1.13,
|
| 87 |
+
"eval_steps_per_second": 1.13,
|
| 88 |
+
"step": 17
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 2.1176470588235294,
|
| 92 |
+
"grad_norm": 2.4443531036376953,
|
| 93 |
+
"learning_rate": 1.5e-06,
|
| 94 |
+
"loss": 2.422,
|
| 95 |
+
"step": 18
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.3529411764705883,
|
| 99 |
+
"grad_norm": 2.2713582515716553,
|
| 100 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 101 |
+
"loss": 2.3315,
|
| 102 |
+
"step": 20
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.588235294117647,
|
| 106 |
+
"grad_norm": 2.6579792499542236,
|
| 107 |
+
"learning_rate": 1.8333333333333333e-06,
|
| 108 |
+
"loss": 2.5587,
|
| 109 |
+
"step": 22
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.8235294117647056,
|
| 113 |
+
"grad_norm": 2.5377650260925293,
|
| 114 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 115 |
+
"loss": 2.46,
|
| 116 |
+
"step": 24
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.9411764705882355,
|
| 120 |
+
"eval_loss": 2.4217796325683594,
|
| 121 |
+
"eval_runtime": 31.8528,
|
| 122 |
+
"eval_samples_per_second": 1.13,
|
| 123 |
+
"eval_steps_per_second": 1.13,
|
| 124 |
+
"step": 25
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 3.0588235294117645,
|
| 128 |
+
"grad_norm": 2.9668917655944824,
|
| 129 |
+
"learning_rate": 2.166666666666667e-06,
|
| 130 |
+
"loss": 2.5217,
|
| 131 |
+
"step": 26
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.2941176470588234,
|
| 135 |
+
"grad_norm": 2.460824728012085,
|
| 136 |
+
"learning_rate": 2.3333333333333336e-06,
|
| 137 |
+
"loss": 2.2339,
|
| 138 |
+
"step": 28
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.5294117647058822,
|
| 142 |
+
"grad_norm": 3.505033493041992,
|
| 143 |
+
"learning_rate": 2.5e-06,
|
| 144 |
+
"loss": 2.4788,
|
| 145 |
+
"step": 30
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.764705882352941,
|
| 149 |
+
"grad_norm": 2.1192469596862793,
|
| 150 |
+
"learning_rate": 2.666666666666667e-06,
|
| 151 |
+
"loss": 2.2748,
|
| 152 |
+
"step": 32
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 4.0,
|
| 156 |
+
"grad_norm": 2.7211689949035645,
|
| 157 |
+
"learning_rate": 2.8333333333333335e-06,
|
| 158 |
+
"loss": 2.4585,
|
| 159 |
+
"step": 34
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 4.0,
|
| 163 |
+
"eval_loss": 2.3277575969696045,
|
| 164 |
+
"eval_runtime": 31.8566,
|
| 165 |
+
"eval_samples_per_second": 1.13,
|
| 166 |
+
"eval_steps_per_second": 1.13,
|
| 167 |
+
"step": 34
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 4.235294117647059,
|
| 171 |
+
"grad_norm": 2.7326908111572266,
|
| 172 |
+
"learning_rate": 3e-06,
|
| 173 |
+
"loss": 2.3561,
|
| 174 |
+
"step": 36
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 4.470588235294118,
|
| 178 |
+
"grad_norm": 2.5006515979766846,
|
| 179 |
+
"learning_rate": 3.1666666666666667e-06,
|
| 180 |
+
"loss": 2.1582,
|
| 181 |
+
"step": 38
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 4.705882352941177,
|
| 185 |
+
"grad_norm": 2.728483200073242,
|
| 186 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 187 |
+
"loss": 2.2463,
|
| 188 |
+
"step": 40
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.9411764705882355,
|
| 192 |
+
"grad_norm": 3.0043787956237793,
|
| 193 |
+
"learning_rate": 3.5e-06,
|
| 194 |
+
"loss": 2.2624,
|
| 195 |
+
"step": 42
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.9411764705882355,
|
| 199 |
+
"eval_loss": 2.191901683807373,
|
| 200 |
+
"eval_runtime": 31.8501,
|
| 201 |
+
"eval_samples_per_second": 1.13,
|
| 202 |
+
"eval_steps_per_second": 1.13,
|
| 203 |
+
"step": 42
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 5.176470588235294,
|
| 207 |
+
"grad_norm": 2.0914738178253174,
|
| 208 |
+
"learning_rate": 3.6666666666666666e-06,
|
| 209 |
+
"loss": 2.0377,
|
| 210 |
+
"step": 44
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 5.411764705882353,
|
| 214 |
+
"grad_norm": 3.9215309619903564,
|
| 215 |
+
"learning_rate": 3.833333333333334e-06,
|
| 216 |
+
"loss": 2.3093,
|
| 217 |
+
"step": 46
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 5.647058823529412,
|
| 221 |
+
"grad_norm": 2.7781357765197754,
|
| 222 |
+
"learning_rate": 4.000000000000001e-06,
|
| 223 |
+
"loss": 2.0491,
|
| 224 |
+
"step": 48
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.882352941176471,
|
| 228 |
+
"grad_norm": 2.7123777866363525,
|
| 229 |
+
"learning_rate": 4.166666666666667e-06,
|
| 230 |
+
"loss": 2.0553,
|
| 231 |
+
"step": 50
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 6.0,
|
| 235 |
+
"eval_loss": 1.9704445600509644,
|
| 236 |
+
"eval_runtime": 31.8507,
|
| 237 |
+
"eval_samples_per_second": 1.13,
|
| 238 |
+
"eval_steps_per_second": 1.13,
|
| 239 |
+
"step": 51
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 6.117647058823529,
|
| 243 |
+
"grad_norm": 3.2437472343444824,
|
| 244 |
+
"learning_rate": 4.333333333333334e-06,
|
| 245 |
+
"loss": 1.9741,
|
| 246 |
+
"step": 52
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 6.352941176470588,
|
| 250 |
+
"grad_norm": 2.4424147605895996,
|
| 251 |
+
"learning_rate": 4.5e-06,
|
| 252 |
+
"loss": 1.8914,
|
| 253 |
+
"step": 54
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 6.588235294117647,
|
| 257 |
+
"grad_norm": 2.605942964553833,
|
| 258 |
+
"learning_rate": 4.666666666666667e-06,
|
| 259 |
+
"loss": 1.8738,
|
| 260 |
+
"step": 56
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.823529411764706,
|
| 264 |
+
"grad_norm": 2.119314432144165,
|
| 265 |
+
"learning_rate": 4.833333333333333e-06,
|
| 266 |
+
"loss": 1.7403,
|
| 267 |
+
"step": 58
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.9411764705882355,
|
| 271 |
+
"eval_loss": 1.7065722942352295,
|
| 272 |
+
"eval_runtime": 31.853,
|
| 273 |
+
"eval_samples_per_second": 1.13,
|
| 274 |
+
"eval_steps_per_second": 1.13,
|
| 275 |
+
"step": 59
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 7.0588235294117645,
|
| 279 |
+
"grad_norm": 1.8015623092651367,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 1.6225,
|
| 282 |
+
"step": 60
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 7.294117647058823,
|
| 286 |
+
"grad_norm": 2.412097215652466,
|
| 287 |
+
"learning_rate": 5.1666666666666675e-06,
|
| 288 |
+
"loss": 1.6174,
|
| 289 |
+
"step": 62
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 7.529411764705882,
|
| 293 |
+
"grad_norm": 2.3760602474212646,
|
| 294 |
+
"learning_rate": 5.333333333333334e-06,
|
| 295 |
+
"loss": 1.5403,
|
| 296 |
+
"step": 64
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.764705882352941,
|
| 300 |
+
"grad_norm": 2.5603954792022705,
|
| 301 |
+
"learning_rate": 5.500000000000001e-06,
|
| 302 |
+
"loss": 1.4378,
|
| 303 |
+
"step": 66
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 8.0,
|
| 307 |
+
"grad_norm": 2.450331211090088,
|
| 308 |
+
"learning_rate": 5.666666666666667e-06,
|
| 309 |
+
"loss": 1.3756,
|
| 310 |
+
"step": 68
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 8.0,
|
| 314 |
+
"eval_loss": 1.3616687059402466,
|
| 315 |
+
"eval_runtime": 31.8467,
|
| 316 |
+
"eval_samples_per_second": 1.13,
|
| 317 |
+
"eval_steps_per_second": 1.13,
|
| 318 |
+
"step": 68
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 8.235294117647058,
|
| 322 |
+
"grad_norm": 1.7808512449264526,
|
| 323 |
+
"learning_rate": 5.833333333333334e-06,
|
| 324 |
+
"loss": 1.2903,
|
| 325 |
+
"step": 70
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 8.470588235294118,
|
| 329 |
+
"grad_norm": 2.280879497528076,
|
| 330 |
+
"learning_rate": 6e-06,
|
| 331 |
+
"loss": 1.2624,
|
| 332 |
+
"step": 72
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.705882352941176,
|
| 336 |
+
"grad_norm": 2.584132671356201,
|
| 337 |
+
"learning_rate": 6.166666666666667e-06,
|
| 338 |
+
"loss": 1.0856,
|
| 339 |
+
"step": 74
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 8.941176470588236,
|
| 343 |
+
"grad_norm": 1.6824058294296265,
|
| 344 |
+
"learning_rate": 6.333333333333333e-06,
|
| 345 |
+
"loss": 1.11,
|
| 346 |
+
"step": 76
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 8.941176470588236,
|
| 350 |
+
"eval_loss": 1.0612744092941284,
|
| 351 |
+
"eval_runtime": 31.847,
|
| 352 |
+
"eval_samples_per_second": 1.13,
|
| 353 |
+
"eval_steps_per_second": 1.13,
|
| 354 |
+
"step": 76
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 9.176470588235293,
|
| 358 |
+
"grad_norm": 1.8082739114761353,
|
| 359 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 360 |
+
"loss": 0.9939,
|
| 361 |
+
"step": 78
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 9.411764705882353,
|
| 365 |
+
"grad_norm": 1.7640292644500732,
|
| 366 |
+
"learning_rate": 6.666666666666667e-06,
|
| 367 |
+
"loss": 0.9378,
|
| 368 |
+
"step": 80
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 9.647058823529411,
|
| 372 |
+
"grad_norm": 1.609857201576233,
|
| 373 |
+
"learning_rate": 6.833333333333334e-06,
|
| 374 |
+
"loss": 0.7986,
|
| 375 |
+
"step": 82
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 9.882352941176471,
|
| 379 |
+
"grad_norm": 1.5625256299972534,
|
| 380 |
+
"learning_rate": 7e-06,
|
| 381 |
+
"loss": 0.7161,
|
| 382 |
+
"step": 84
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 10.0,
|
| 386 |
+
"eval_loss": 0.7772443890571594,
|
| 387 |
+
"eval_runtime": 31.8458,
|
| 388 |
+
"eval_samples_per_second": 1.13,
|
| 389 |
+
"eval_steps_per_second": 1.13,
|
| 390 |
+
"step": 85
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 10.117647058823529,
|
| 394 |
+
"grad_norm": 1.2514125108718872,
|
| 395 |
+
"learning_rate": 7.166666666666667e-06,
|
| 396 |
+
"loss": 0.7059,
|
| 397 |
+
"step": 86
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 10.352941176470589,
|
| 401 |
+
"grad_norm": 0.9835730195045471,
|
| 402 |
+
"learning_rate": 7.333333333333333e-06,
|
| 403 |
+
"loss": 0.5005,
|
| 404 |
+
"step": 88
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 10.588235294117647,
|
| 408 |
+
"grad_norm": 0.8222648501396179,
|
| 409 |
+
"learning_rate": 7.500000000000001e-06,
|
| 410 |
+
"loss": 0.7141,
|
| 411 |
+
"step": 90
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 10.823529411764707,
|
| 415 |
+
"grad_norm": 1.2220810651779175,
|
| 416 |
+
"learning_rate": 7.666666666666667e-06,
|
| 417 |
+
"loss": 0.7609,
|
| 418 |
+
"step": 92
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 10.941176470588236,
|
| 422 |
+
"eval_loss": 0.6786696910858154,
|
| 423 |
+
"eval_runtime": 31.8483,
|
| 424 |
+
"eval_samples_per_second": 1.13,
|
| 425 |
+
"eval_steps_per_second": 1.13,
|
| 426 |
+
"step": 93
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 11.058823529411764,
|
| 430 |
+
"grad_norm": 0.6009970307350159,
|
| 431 |
+
"learning_rate": 7.833333333333333e-06,
|
| 432 |
+
"loss": 0.6874,
|
| 433 |
+
"step": 94
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 11.294117647058824,
|
| 437 |
+
"grad_norm": 0.7795246243476868,
|
| 438 |
+
"learning_rate": 8.000000000000001e-06,
|
| 439 |
+
"loss": 0.7462,
|
| 440 |
+
"step": 96
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 11.529411764705882,
|
| 444 |
+
"grad_norm": 0.6197288036346436,
|
| 445 |
+
"learning_rate": 8.166666666666668e-06,
|
| 446 |
+
"loss": 0.6408,
|
| 447 |
+
"step": 98
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 11.764705882352942,
|
| 451 |
+
"grad_norm": 0.6273565292358398,
|
| 452 |
+
"learning_rate": 8.333333333333334e-06,
|
| 453 |
+
"loss": 0.4579,
|
| 454 |
+
"step": 100
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 12.0,
|
| 458 |
+
"grad_norm": 0.45476672053337097,
|
| 459 |
+
"learning_rate": 8.5e-06,
|
| 460 |
+
"loss": 0.4358,
|
| 461 |
+
"step": 102
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 12.0,
|
| 465 |
+
"eval_loss": 0.6182084083557129,
|
| 466 |
+
"eval_runtime": 31.8454,
|
| 467 |
+
"eval_samples_per_second": 1.13,
|
| 468 |
+
"eval_steps_per_second": 1.13,
|
| 469 |
+
"step": 102
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 12.235294117647058,
|
| 473 |
+
"grad_norm": 0.7051817774772644,
|
| 474 |
+
"learning_rate": 8.666666666666668e-06,
|
| 475 |
+
"loss": 0.6431,
|
| 476 |
+
"step": 104
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 12.470588235294118,
|
| 480 |
+
"grad_norm": 0.5281614065170288,
|
| 481 |
+
"learning_rate": 8.833333333333334e-06,
|
| 482 |
+
"loss": 0.518,
|
| 483 |
+
"step": 106
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 12.705882352941176,
|
| 487 |
+
"grad_norm": 0.5485586524009705,
|
| 488 |
+
"learning_rate": 9e-06,
|
| 489 |
+
"loss": 0.4787,
|
| 490 |
+
"step": 108
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 12.941176470588236,
|
| 494 |
+
"grad_norm": 0.9739407300949097,
|
| 495 |
+
"learning_rate": 9.166666666666666e-06,
|
| 496 |
+
"loss": 0.4774,
|
| 497 |
+
"step": 110
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 12.941176470588236,
|
| 501 |
+
"eval_loss": 0.5912042260169983,
|
| 502 |
+
"eval_runtime": 31.8561,
|
| 503 |
+
"eval_samples_per_second": 1.13,
|
| 504 |
+
"eval_steps_per_second": 1.13,
|
| 505 |
+
"step": 110
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 13.176470588235293,
|
| 509 |
+
"grad_norm": 0.3852427899837494,
|
| 510 |
+
"learning_rate": 9.333333333333334e-06,
|
| 511 |
+
"loss": 0.476,
|
| 512 |
+
"step": 112
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 13.411764705882353,
|
| 516 |
+
"grad_norm": 0.3773607909679413,
|
| 517 |
+
"learning_rate": 9.5e-06,
|
| 518 |
+
"loss": 0.4432,
|
| 519 |
+
"step": 114
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 13.647058823529411,
|
| 523 |
+
"grad_norm": 0.5833092331886292,
|
| 524 |
+
"learning_rate": 9.666666666666667e-06,
|
| 525 |
+
"loss": 0.4677,
|
| 526 |
+
"step": 116
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 13.882352941176471,
|
| 530 |
+
"grad_norm": 0.525675356388092,
|
| 531 |
+
"learning_rate": 9.833333333333333e-06,
|
| 532 |
+
"loss": 0.5569,
|
| 533 |
+
"step": 118
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 14.0,
|
| 537 |
+
"eval_loss": 0.5746464133262634,
|
| 538 |
+
"eval_runtime": 31.8497,
|
| 539 |
+
"eval_samples_per_second": 1.13,
|
| 540 |
+
"eval_steps_per_second": 1.13,
|
| 541 |
+
"step": 119
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 14.117647058823529,
|
| 545 |
+
"grad_norm": 0.46034735441207886,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 0.6595,
|
| 548 |
+
"step": 120
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 14.352941176470589,
|
| 552 |
+
"grad_norm": 0.9925610423088074,
|
| 553 |
+
"learning_rate": 9.999915384288723e-06,
|
| 554 |
+
"loss": 0.5066,
|
| 555 |
+
"step": 122
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 14.588235294117647,
|
| 559 |
+
"grad_norm": 0.3283170163631439,
|
| 560 |
+
"learning_rate": 9.999661540018812e-06,
|
| 561 |
+
"loss": 0.5017,
|
| 562 |
+
"step": 124
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 14.823529411764707,
|
| 566 |
+
"grad_norm": 0.4828791320323944,
|
| 567 |
+
"learning_rate": 9.999238475781957e-06,
|
| 568 |
+
"loss": 0.427,
|
| 569 |
+
"step": 126
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 14.941176470588236,
|
| 573 |
+
"eval_loss": 0.5487004518508911,
|
| 574 |
+
"eval_runtime": 31.8495,
|
| 575 |
+
"eval_samples_per_second": 1.13,
|
| 576 |
+
"eval_steps_per_second": 1.13,
|
| 577 |
+
"step": 127
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 15.058823529411764,
|
| 581 |
+
"grad_norm": 0.44912636280059814,
|
| 582 |
+
"learning_rate": 9.99864620589731e-06,
|
| 583 |
+
"loss": 0.4705,
|
| 584 |
+
"step": 128
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 15.294117647058824,
|
| 588 |
+
"grad_norm": 0.43985074758529663,
|
| 589 |
+
"learning_rate": 9.997884750411004e-06,
|
| 590 |
+
"loss": 0.3865,
|
| 591 |
+
"step": 130
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 15.529411764705882,
|
| 595 |
+
"grad_norm": 1.4928995370864868,
|
| 596 |
+
"learning_rate": 9.99695413509548e-06,
|
| 597 |
+
"loss": 0.483,
|
| 598 |
+
"step": 132
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 15.764705882352942,
|
| 602 |
+
"grad_norm": 0.3390346169471741,
|
| 603 |
+
"learning_rate": 9.995854391448607e-06,
|
| 604 |
+
"loss": 0.5068,
|
| 605 |
+
"step": 134
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 16.0,
|
| 609 |
+
"grad_norm": 0.33858582377433777,
|
| 610 |
+
"learning_rate": 9.994585556692624e-06,
|
| 611 |
+
"loss": 0.4672,
|
| 612 |
+
"step": 136
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 16.0,
|
| 616 |
+
"eval_loss": 0.533892035484314,
|
| 617 |
+
"eval_runtime": 31.8489,
|
| 618 |
+
"eval_samples_per_second": 1.13,
|
| 619 |
+
"eval_steps_per_second": 1.13,
|
| 620 |
+
"step": 136
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 16.235294117647058,
|
| 624 |
+
"grad_norm": 0.3130759298801422,
|
| 625 |
+
"learning_rate": 9.993147673772869e-06,
|
| 626 |
+
"loss": 0.3805,
|
| 627 |
+
"step": 138
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 16.470588235294116,
|
| 631 |
+
"grad_norm": 0.32997503876686096,
|
| 632 |
+
"learning_rate": 9.991540791356342e-06,
|
| 633 |
+
"loss": 0.6362,
|
| 634 |
+
"step": 140
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 16.705882352941178,
|
| 638 |
+
"grad_norm": 1.398141622543335,
|
| 639 |
+
"learning_rate": 9.989764963830038e-06,
|
| 640 |
+
"loss": 0.4943,
|
| 641 |
+
"step": 142
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 16.941176470588236,
|
| 645 |
+
"grad_norm": 0.19720911979675293,
|
| 646 |
+
"learning_rate": 9.987820251299121e-06,
|
| 647 |
+
"loss": 0.3495,
|
| 648 |
+
"step": 144
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 16.941176470588236,
|
| 652 |
+
"eval_loss": 0.552502453327179,
|
| 653 |
+
"eval_runtime": 31.8645,
|
| 654 |
+
"eval_samples_per_second": 1.13,
|
| 655 |
+
"eval_steps_per_second": 1.13,
|
| 656 |
+
"step": 144
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 17.176470588235293,
|
| 660 |
+
"grad_norm": 0.21927881240844727,
|
| 661 |
+
"learning_rate": 9.985706719584888e-06,
|
| 662 |
+
"loss": 0.4342,
|
| 663 |
+
"step": 146
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 17.41176470588235,
|
| 667 |
+
"grad_norm": 0.2400996834039688,
|
| 668 |
+
"learning_rate": 9.98342444022253e-06,
|
| 669 |
+
"loss": 0.4801,
|
| 670 |
+
"step": 148
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 17.647058823529413,
|
| 674 |
+
"grad_norm": 0.2719495892524719,
|
| 675 |
+
"learning_rate": 9.980973490458728e-06,
|
| 676 |
+
"loss": 0.4199,
|
| 677 |
+
"step": 150
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 17.88235294117647,
|
| 681 |
+
"grad_norm": 0.18624863028526306,
|
| 682 |
+
"learning_rate": 9.978353953249023e-06,
|
| 683 |
+
"loss": 0.4731,
|
| 684 |
+
"step": 152
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 18.0,
|
| 688 |
+
"eval_loss": 0.5323101878166199,
|
| 689 |
+
"eval_runtime": 31.8439,
|
| 690 |
+
"eval_samples_per_second": 1.131,
|
| 691 |
+
"eval_steps_per_second": 1.131,
|
| 692 |
+
"step": 153
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 18.11764705882353,
|
| 696 |
+
"grad_norm": 0.18268544971942902,
|
| 697 |
+
"learning_rate": 9.975565917255017e-06,
|
| 698 |
+
"loss": 0.3459,
|
| 699 |
+
"step": 154
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 18.352941176470587,
|
| 703 |
+
"grad_norm": 0.30436864495277405,
|
| 704 |
+
"learning_rate": 9.972609476841368e-06,
|
| 705 |
+
"loss": 0.58,
|
| 706 |
+
"step": 156
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 18.58823529411765,
|
| 710 |
+
"grad_norm": 0.2947259843349457,
|
| 711 |
+
"learning_rate": 9.9694847320726e-06,
|
| 712 |
+
"loss": 0.4085,
|
| 713 |
+
"step": 158
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 18.823529411764707,
|
| 717 |
+
"grad_norm": 0.24882009625434875,
|
| 718 |
+
"learning_rate": 9.966191788709716e-06,
|
| 719 |
+
"loss": 0.3913,
|
| 720 |
+
"step": 160
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 18.941176470588236,
|
| 724 |
+
"eval_loss": 0.5243125557899475,
|
| 725 |
+
"eval_runtime": 31.8447,
|
| 726 |
+
"eval_samples_per_second": 1.13,
|
| 727 |
+
"eval_steps_per_second": 1.13,
|
| 728 |
+
"step": 161
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 19.058823529411764,
|
| 732 |
+
"grad_norm": 0.31586578488349915,
|
| 733 |
+
"learning_rate": 9.962730758206612e-06,
|
| 734 |
+
"loss": 0.5022,
|
| 735 |
+
"step": 162
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 19.294117647058822,
|
| 739 |
+
"grad_norm": 0.21429723501205444,
|
| 740 |
+
"learning_rate": 9.959101757706308e-06,
|
| 741 |
+
"loss": 0.3682,
|
| 742 |
+
"step": 164
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 19.529411764705884,
|
| 746 |
+
"grad_norm": 0.29345616698265076,
|
| 747 |
+
"learning_rate": 9.955304910036993e-06,
|
| 748 |
+
"loss": 0.4185,
|
| 749 |
+
"step": 166
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 19.764705882352942,
|
| 753 |
+
"grad_norm": 0.24088309705257416,
|
| 754 |
+
"learning_rate": 9.951340343707852e-06,
|
| 755 |
+
"loss": 0.4027,
|
| 756 |
+
"step": 168
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 20.0,
|
| 760 |
+
"grad_norm": 0.3225376307964325,
|
| 761 |
+
"learning_rate": 9.947208192904722e-06,
|
| 762 |
+
"loss": 0.5624,
|
| 763 |
+
"step": 170
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 20.0,
|
| 767 |
+
"eval_loss": 0.5252914428710938,
|
| 768 |
+
"eval_runtime": 31.8447,
|
| 769 |
+
"eval_samples_per_second": 1.13,
|
| 770 |
+
"eval_steps_per_second": 1.13,
|
| 771 |
+
"step": 170
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 20.235294117647058,
|
| 775 |
+
"grad_norm": 0.24431054294109344,
|
| 776 |
+
"learning_rate": 9.942908597485558e-06,
|
| 777 |
+
"loss": 0.3565,
|
| 778 |
+
"step": 172
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 20.470588235294116,
|
| 782 |
+
"grad_norm": 0.30302417278289795,
|
| 783 |
+
"learning_rate": 9.938441702975689e-06,
|
| 784 |
+
"loss": 0.4944,
|
| 785 |
+
"step": 174
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 20.705882352941178,
|
| 789 |
+
"grad_norm": 0.25407275557518005,
|
| 790 |
+
"learning_rate": 9.933807660562898e-06,
|
| 791 |
+
"loss": 0.3397,
|
| 792 |
+
"step": 176
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 20.941176470588236,
|
| 796 |
+
"grad_norm": 0.3202497363090515,
|
| 797 |
+
"learning_rate": 9.929006627092298e-06,
|
| 798 |
+
"loss": 0.4684,
|
| 799 |
+
"step": 178
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 20.941176470588236,
|
| 803 |
+
"eval_loss": 0.5222189426422119,
|
| 804 |
+
"eval_runtime": 31.8492,
|
| 805 |
+
"eval_samples_per_second": 1.13,
|
| 806 |
+
"eval_steps_per_second": 1.13,
|
| 807 |
+
"step": 178
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 21.176470588235293,
|
| 811 |
+
"grad_norm": 0.2713940441608429,
|
| 812 |
+
"learning_rate": 9.924038765061042e-06,
|
| 813 |
+
"loss": 0.5016,
|
| 814 |
+
"step": 180
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 21.41176470588235,
|
| 818 |
+
"grad_norm": 0.2259199619293213,
|
| 819 |
+
"learning_rate": 9.918904242612794e-06,
|
| 820 |
+
"loss": 0.5315,
|
| 821 |
+
"step": 182
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 21.647058823529413,
|
| 825 |
+
"grad_norm": 0.2454088032245636,
|
| 826 |
+
"learning_rate": 9.913603233532067e-06,
|
| 827 |
+
"loss": 0.474,
|
| 828 |
+
"step": 184
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 21.88235294117647,
|
| 832 |
+
"grad_norm": 0.18309438228607178,
|
| 833 |
+
"learning_rate": 9.908135917238321e-06,
|
| 834 |
+
"loss": 0.3029,
|
| 835 |
+
"step": 186
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 22.0,
|
| 839 |
+
"eval_loss": 0.5099583864212036,
|
| 840 |
+
"eval_runtime": 31.8371,
|
| 841 |
+
"eval_samples_per_second": 1.131,
|
| 842 |
+
"eval_steps_per_second": 1.131,
|
| 843 |
+
"step": 187
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 22.11764705882353,
|
| 847 |
+
"grad_norm": 0.3067823648452759,
|
| 848 |
+
"learning_rate": 9.902502478779897e-06,
|
| 849 |
+
"loss": 0.3838,
|
| 850 |
+
"step": 188
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 22.352941176470587,
|
| 854 |
+
"grad_norm": 0.2594364881515503,
|
| 855 |
+
"learning_rate": 9.896703108827758e-06,
|
| 856 |
+
"loss": 0.48,
|
| 857 |
+
"step": 190
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 22.58823529411765,
|
| 861 |
+
"grad_norm": 0.27693891525268555,
|
| 862 |
+
"learning_rate": 9.890738003669029e-06,
|
| 863 |
+
"loss": 0.4363,
|
| 864 |
+
"step": 192
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 22.823529411764707,
|
| 868 |
+
"grad_norm": 0.304267019033432,
|
| 869 |
+
"learning_rate": 9.884607365200355e-06,
|
| 870 |
+
"loss": 0.3522,
|
| 871 |
+
"step": 194
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 22.941176470588236,
|
| 875 |
+
"eval_loss": 0.5085048079490662,
|
| 876 |
+
"eval_runtime": 31.8502,
|
| 877 |
+
"eval_samples_per_second": 1.13,
|
| 878 |
+
"eval_steps_per_second": 1.13,
|
| 879 |
+
"step": 195
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 23.058823529411764,
|
| 883 |
+
"grad_norm": 0.17085126042366028,
|
| 884 |
+
"learning_rate": 9.878311400921072e-06,
|
| 885 |
+
"loss": 0.2722,
|
| 886 |
+
"step": 196
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 23.294117647058822,
|
| 890 |
+
"grad_norm": 0.4169885218143463,
|
| 891 |
+
"learning_rate": 9.871850323926178e-06,
|
| 892 |
+
"loss": 0.4118,
|
| 893 |
+
"step": 198
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 23.529411764705884,
|
| 897 |
+
"grad_norm": 0.25028589367866516,
|
| 898 |
+
"learning_rate": 9.86522435289912e-06,
|
| 899 |
+
"loss": 0.3777,
|
| 900 |
+
"step": 200
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 23.764705882352942,
|
| 904 |
+
"grad_norm": 0.3203675150871277,
|
| 905 |
+
"learning_rate": 9.858433712104403e-06,
|
| 906 |
+
"loss": 0.5195,
|
| 907 |
+
"step": 202
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 24.0,
|
| 911 |
+
"grad_norm": 0.3470858931541443,
|
| 912 |
+
"learning_rate": 9.851478631379982e-06,
|
| 913 |
+
"loss": 0.3855,
|
| 914 |
+
"step": 204
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 24.0,
|
| 918 |
+
"eval_loss": 0.49713531136512756,
|
| 919 |
+
"eval_runtime": 31.8498,
|
| 920 |
+
"eval_samples_per_second": 1.13,
|
| 921 |
+
"eval_steps_per_second": 1.13,
|
| 922 |
+
"step": 204
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 24.235294117647058,
|
| 926 |
+
"grad_norm": 0.3214435875415802,
|
| 927 |
+
"learning_rate": 9.844359346129504e-06,
|
| 928 |
+
"loss": 0.476,
|
| 929 |
+
"step": 206
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 24.470588235294116,
|
| 933 |
+
"grad_norm": 0.3332065939903259,
|
| 934 |
+
"learning_rate": 9.83707609731432e-06,
|
| 935 |
+
"loss": 0.4383,
|
| 936 |
+
"step": 208
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 24.705882352941178,
|
| 940 |
+
"grad_norm": 0.22289708256721497,
|
| 941 |
+
"learning_rate": 9.829629131445342e-06,
|
| 942 |
+
"loss": 0.3256,
|
| 943 |
+
"step": 210
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 24.941176470588236,
|
| 947 |
+
"grad_norm": 0.25234198570251465,
|
| 948 |
+
"learning_rate": 9.822018700574696e-06,
|
| 949 |
+
"loss": 0.317,
|
| 950 |
+
"step": 212
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 24.941176470588236,
|
| 954 |
+
"eval_loss": 0.504925549030304,
|
| 955 |
+
"eval_runtime": 31.8487,
|
| 956 |
+
"eval_samples_per_second": 1.13,
|
| 957 |
+
"eval_steps_per_second": 1.13,
|
| 958 |
+
"step": 212
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 25.176470588235293,
|
| 962 |
+
"grad_norm": 0.3678690493106842,
|
| 963 |
+
"learning_rate": 9.81424506228719e-06,
|
| 964 |
+
"loss": 0.4398,
|
| 965 |
+
"step": 214
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 25.41176470588235,
|
| 969 |
+
"grad_norm": 0.31491589546203613,
|
| 970 |
+
"learning_rate": 9.806308479691595e-06,
|
| 971 |
+
"loss": 0.3335,
|
| 972 |
+
"step": 216
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 25.647058823529413,
|
| 976 |
+
"grad_norm": 0.3429551124572754,
|
| 977 |
+
"learning_rate": 9.798209221411748e-06,
|
| 978 |
+
"loss": 0.4467,
|
| 979 |
+
"step": 218
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 25.88235294117647,
|
| 983 |
+
"grad_norm": 0.30439844727516174,
|
| 984 |
+
"learning_rate": 9.789947561577445e-06,
|
| 985 |
+
"loss": 0.338,
|
| 986 |
+
"step": 220
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 26.0,
|
| 990 |
+
"eval_loss": 0.5015532970428467,
|
| 991 |
+
"eval_runtime": 31.8415,
|
| 992 |
+
"eval_samples_per_second": 1.131,
|
| 993 |
+
"eval_steps_per_second": 1.131,
|
| 994 |
+
"step": 221
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 26.11764705882353,
|
| 998 |
+
"grad_norm": 0.2530220150947571,
|
| 999 |
+
"learning_rate": 9.781523779815178e-06,
|
| 1000 |
+
"loss": 0.3514,
|
| 1001 |
+
"step": 222
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 26.352941176470587,
|
| 1005 |
+
"grad_norm": 0.2835986018180847,
|
| 1006 |
+
"learning_rate": 9.77293816123866e-06,
|
| 1007 |
+
"loss": 0.3022,
|
| 1008 |
+
"step": 224
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 26.58823529411765,
|
| 1012 |
+
"grad_norm": 0.3710886538028717,
|
| 1013 |
+
"learning_rate": 9.764190996439181e-06,
|
| 1014 |
+
"loss": 0.4567,
|
| 1015 |
+
"step": 226
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 26.823529411764707,
|
| 1019 |
+
"grad_norm": 0.3602848947048187,
|
| 1020 |
+
"learning_rate": 9.755282581475769e-06,
|
| 1021 |
+
"loss": 0.391,
|
| 1022 |
+
"step": 228
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 26.941176470588236,
|
| 1026 |
+
"eval_loss": 0.49422508478164673,
|
| 1027 |
+
"eval_runtime": 31.8515,
|
| 1028 |
+
"eval_samples_per_second": 1.13,
|
| 1029 |
+
"eval_steps_per_second": 1.13,
|
| 1030 |
+
"step": 229
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 27.058823529411764,
|
| 1034 |
+
"grad_norm": 0.3394179940223694,
|
| 1035 |
+
"learning_rate": 9.74621321786517e-06,
|
| 1036 |
+
"loss": 0.3986,
|
| 1037 |
+
"step": 230
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 27.294117647058822,
|
| 1041 |
+
"grad_norm": 0.4131983518600464,
|
| 1042 |
+
"learning_rate": 9.736983212571646e-06,
|
| 1043 |
+
"loss": 0.359,
|
| 1044 |
+
"step": 232
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 27.529411764705884,
|
| 1048 |
+
"grad_norm": 0.35208216309547424,
|
| 1049 |
+
"learning_rate": 9.727592877996585e-06,
|
| 1050 |
+
"loss": 0.2981,
|
| 1051 |
+
"step": 234
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 27.764705882352942,
|
| 1055 |
+
"grad_norm": 0.42573440074920654,
|
| 1056 |
+
"learning_rate": 9.718042531967918e-06,
|
| 1057 |
+
"loss": 0.4258,
|
| 1058 |
+
"step": 236
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 28.0,
|
| 1062 |
+
"grad_norm": 0.3431633412837982,
|
| 1063 |
+
"learning_rate": 9.708332497729378e-06,
|
| 1064 |
+
"loss": 0.3964,
|
| 1065 |
+
"step": 238
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 28.0,
|
| 1069 |
+
"eval_loss": 0.5010454058647156,
|
| 1070 |
+
"eval_runtime": 31.8416,
|
| 1071 |
+
"eval_samples_per_second": 1.131,
|
| 1072 |
+
"eval_steps_per_second": 1.131,
|
| 1073 |
+
"step": 238
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 28.235294117647058,
|
| 1077 |
+
"grad_norm": 0.31860944628715515,
|
| 1078 |
+
"learning_rate": 9.698463103929542e-06,
|
| 1079 |
+
"loss": 0.2962,
|
| 1080 |
+
"step": 240
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 28.470588235294116,
|
| 1084 |
+
"grad_norm": 0.34731313586235046,
|
| 1085 |
+
"learning_rate": 9.688434684610725e-06,
|
| 1086 |
+
"loss": 0.3673,
|
| 1087 |
+
"step": 242
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 28.705882352941178,
|
| 1091 |
+
"grad_norm": 0.36615273356437683,
|
| 1092 |
+
"learning_rate": 9.678247579197658e-06,
|
| 1093 |
+
"loss": 0.5272,
|
| 1094 |
+
"step": 244
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 28.941176470588236,
|
| 1098 |
+
"grad_norm": 0.2973138689994812,
|
| 1099 |
+
"learning_rate": 9.667902132486009e-06,
|
| 1100 |
+
"loss": 0.2951,
|
| 1101 |
+
"step": 246
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 28.941176470588236,
|
| 1105 |
+
"eval_loss": 0.509841799736023,
|
| 1106 |
+
"eval_runtime": 31.843,
|
| 1107 |
+
"eval_samples_per_second": 1.131,
|
| 1108 |
+
"eval_steps_per_second": 1.131,
|
| 1109 |
+
"step": 246
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 29.176470588235293,
|
| 1113 |
+
"grad_norm": 0.3237411379814148,
|
| 1114 |
+
"learning_rate": 9.657398694630713e-06,
|
| 1115 |
+
"loss": 0.2387,
|
| 1116 |
+
"step": 248
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 29.41176470588235,
|
| 1120 |
+
"grad_norm": 0.31316205859184265,
|
| 1121 |
+
"learning_rate": 9.646737621134112e-06,
|
| 1122 |
+
"loss": 0.3701,
|
| 1123 |
+
"step": 250
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 29.647058823529413,
|
| 1127 |
+
"grad_norm": 0.41230136156082153,
|
| 1128 |
+
"learning_rate": 9.635919272833938e-06,
|
| 1129 |
+
"loss": 0.4157,
|
| 1130 |
+
"step": 252
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 29.88235294117647,
|
| 1134 |
+
"grad_norm": 0.4581688940525055,
|
| 1135 |
+
"learning_rate": 9.62494401589108e-06,
|
| 1136 |
+
"loss": 0.4021,
|
| 1137 |
+
"step": 254
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 30.0,
|
| 1141 |
+
"eval_loss": 0.5068490505218506,
|
| 1142 |
+
"eval_runtime": 31.8451,
|
| 1143 |
+
"eval_samples_per_second": 1.13,
|
| 1144 |
+
"eval_steps_per_second": 1.13,
|
| 1145 |
+
"step": 255
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 30.11764705882353,
|
| 1149 |
+
"grad_norm": 0.3983169198036194,
|
| 1150 |
+
"learning_rate": 9.613812221777212e-06,
|
| 1151 |
+
"loss": 0.3278,
|
| 1152 |
+
"step": 256
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 30.352941176470587,
|
| 1156 |
+
"grad_norm": 0.26158151030540466,
|
| 1157 |
+
"learning_rate": 9.602524267262202e-06,
|
| 1158 |
+
"loss": 0.1625,
|
| 1159 |
+
"step": 258
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 30.58823529411765,
|
| 1163 |
+
"grad_norm": 0.4597890377044678,
|
| 1164 |
+
"learning_rate": 9.591080534401371e-06,
|
| 1165 |
+
"loss": 0.4323,
|
| 1166 |
+
"step": 260
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 30.823529411764707,
|
| 1170 |
+
"grad_norm": 0.4650469422340393,
|
| 1171 |
+
"learning_rate": 9.579481410522556e-06,
|
| 1172 |
+
"loss": 0.4021,
|
| 1173 |
+
"step": 262
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 30.941176470588236,
|
| 1177 |
+
"eval_loss": 0.507030189037323,
|
| 1178 |
+
"eval_runtime": 31.8372,
|
| 1179 |
+
"eval_samples_per_second": 1.131,
|
| 1180 |
+
"eval_steps_per_second": 1.131,
|
| 1181 |
+
"step": 263
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 31.058823529411764,
|
| 1185 |
+
"grad_norm": 0.41834914684295654,
|
| 1186 |
+
"learning_rate": 9.567727288213005e-06,
|
| 1187 |
+
"loss": 0.3066,
|
| 1188 |
+
"step": 264
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 31.294117647058822,
|
| 1192 |
+
"grad_norm": 0.40843620896339417,
|
| 1193 |
+
"learning_rate": 9.555818565306086e-06,
|
| 1194 |
+
"loss": 0.288,
|
| 1195 |
+
"step": 266
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 31.529411764705884,
|
| 1199 |
+
"grad_norm": 0.3444281816482544,
|
| 1200 |
+
"learning_rate": 9.543755644867823e-06,
|
| 1201 |
+
"loss": 0.3391,
|
| 1202 |
+
"step": 268
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 31.764705882352942,
|
| 1206 |
+
"grad_norm": 0.42617908120155334,
|
| 1207 |
+
"learning_rate": 9.531538935183252e-06,
|
| 1208 |
+
"loss": 0.396,
|
| 1209 |
+
"step": 270
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 32.0,
|
| 1213 |
+
"grad_norm": 0.47771233320236206,
|
| 1214 |
+
"learning_rate": 9.519168849742603e-06,
|
| 1215 |
+
"loss": 0.3456,
|
| 1216 |
+
"step": 272
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 32.0,
|
| 1220 |
+
"eval_loss": 0.5024969577789307,
|
| 1221 |
+
"eval_runtime": 31.841,
|
| 1222 |
+
"eval_samples_per_second": 1.131,
|
| 1223 |
+
"eval_steps_per_second": 1.131,
|
| 1224 |
+
"step": 272
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 32.23529411764706,
|
| 1228 |
+
"grad_norm": 0.5327393412590027,
|
| 1229 |
+
"learning_rate": 9.506645807227311e-06,
|
| 1230 |
+
"loss": 0.3042,
|
| 1231 |
+
"step": 274
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 32.470588235294116,
|
| 1235 |
+
"grad_norm": 0.35559046268463135,
|
| 1236 |
+
"learning_rate": 9.493970231495836e-06,
|
| 1237 |
+
"loss": 0.386,
|
| 1238 |
+
"step": 276
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 32.705882352941174,
|
| 1242 |
+
"grad_norm": 0.3480581045150757,
|
| 1243 |
+
"learning_rate": 9.481142551569318e-06,
|
| 1244 |
+
"loss": 0.2265,
|
| 1245 |
+
"step": 278
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 32.94117647058823,
|
| 1249 |
+
"grad_norm": 0.5482520461082458,
|
| 1250 |
+
"learning_rate": 9.468163201617063e-06,
|
| 1251 |
+
"loss": 0.4431,
|
| 1252 |
+
"step": 280
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 32.94117647058823,
|
| 1256 |
+
"eval_loss": 0.5049856901168823,
|
| 1257 |
+
"eval_runtime": 31.8427,
|
| 1258 |
+
"eval_samples_per_second": 1.131,
|
| 1259 |
+
"eval_steps_per_second": 1.131,
|
| 1260 |
+
"step": 280
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 33.1764705882353,
|
| 1264 |
+
"grad_norm": 0.3856636583805084,
|
| 1265 |
+
"learning_rate": 9.45503262094184e-06,
|
| 1266 |
+
"loss": 0.1491,
|
| 1267 |
+
"step": 282
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 33.411764705882355,
|
| 1271 |
+
"grad_norm": 0.5228236317634583,
|
| 1272 |
+
"learning_rate": 9.441751253965022e-06,
|
| 1273 |
+
"loss": 0.3196,
|
| 1274 |
+
"step": 284
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 33.64705882352941,
|
| 1278 |
+
"grad_norm": 0.38300809264183044,
|
| 1279 |
+
"learning_rate": 9.428319550211531e-06,
|
| 1280 |
+
"loss": 0.2612,
|
| 1281 |
+
"step": 286
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 33.88235294117647,
|
| 1285 |
+
"grad_norm": 0.5598320364952087,
|
| 1286 |
+
"learning_rate": 9.414737964294636e-06,
|
| 1287 |
+
"loss": 0.4131,
|
| 1288 |
+
"step": 288
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 34.0,
|
| 1292 |
+
"eval_loss": 0.5093716382980347,
|
| 1293 |
+
"eval_runtime": 31.8401,
|
| 1294 |
+
"eval_samples_per_second": 1.131,
|
| 1295 |
+
"eval_steps_per_second": 1.131,
|
| 1296 |
+
"step": 289
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 34.0,
|
| 1300 |
+
"step": 289,
|
| 1301 |
+
"total_flos": 3.6265283397720474e+17,
|
| 1302 |
+
"train_loss": 0.8757064017663777,
|
| 1303 |
+
"train_runtime": 13296.8495,
|
| 1304 |
+
"train_samples_per_second": 1.534,
|
| 1305 |
+
"train_steps_per_second": 0.09
|
| 1306 |
+
}
|
| 1307 |
+
],
|
| 1308 |
+
"logging_steps": 2,
|
| 1309 |
+
"max_steps": 1200,
|
| 1310 |
+
"num_input_tokens_seen": 0,
|
| 1311 |
+
"num_train_epochs": 150,
|
| 1312 |
+
"save_steps": 25,
|
| 1313 |
+
"stateful_callbacks": {
|
| 1314 |
+
"EarlyStoppingCallback": {
|
| 1315 |
+
"args": {
|
| 1316 |
+
"early_stopping_patience": 7,
|
| 1317 |
+
"early_stopping_threshold": 0.0
|
| 1318 |
+
},
|
| 1319 |
+
"attributes": {
|
| 1320 |
+
"early_stopping_patience_counter": 0
|
| 1321 |
+
}
|
| 1322 |
+
},
|
| 1323 |
+
"TrainerControl": {
|
| 1324 |
+
"args": {
|
| 1325 |
+
"should_epoch_stop": false,
|
| 1326 |
+
"should_evaluate": false,
|
| 1327 |
+
"should_log": false,
|
| 1328 |
+
"should_save": true,
|
| 1329 |
+
"should_training_stop": true
|
| 1330 |
+
},
|
| 1331 |
+
"attributes": {}
|
| 1332 |
+
}
|
| 1333 |
+
},
|
| 1334 |
+
"total_flos": 3.6265283397720474e+17,
|
| 1335 |
+
"train_batch_size": 1,
|
| 1336 |
+
"trial_name": null,
|
| 1337 |
+
"trial_params": null
|
| 1338 |
+
}
|