Model save
Browse files- README.md +90 -0
- all_results.json +9 -0
- generation_config.json +7 -0
- model.safetensors +1 -1
- train_results.json +9 -0
- trainer_state.json +595 -0
README.md
ADDED
|
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: hZzy/qwen2.5-0.5b-sft-news-IFT
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- expo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
model-index:
|
| 9 |
+
- name: qwen2.5-0.5b-expo-DPO-ES-10
|
| 10 |
+
results: []
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
+
|
| 16 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/zhiyuzha-university-of-florida/huggingface/runs/8420vo52)
|
| 17 |
+
# qwen2.5-0.5b-expo-DPO-ES-10
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [hZzy/qwen2.5-0.5b-sft-news-IFT](https://huggingface.co/hZzy/qwen2.5-0.5b-sft-news-IFT) on an unknown dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 21.3539
|
| 22 |
+
- Logps: -79.7115
|
| 23 |
+
- Logits: -0.5475
|
| 24 |
+
- Objective: 20.4532
|
| 25 |
+
- Dpo Loss: 20.4532
|
| 26 |
+
- Regularize: 20.4532
|
| 27 |
+
- Ranking Simple: 0.5362
|
| 28 |
+
- Ranking Idealized: 0.5212
|
| 29 |
+
- Ranking Idealized Expo: 0.5212
|
| 30 |
+
- Wo Beta: 6.6867
|
| 31 |
+
|
| 32 |
+
## Model description
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Intended uses & limitations
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Training and evaluation data
|
| 41 |
+
|
| 42 |
+
More information needed
|
| 43 |
+
|
| 44 |
+
## Training procedure
|
| 45 |
+
|
| 46 |
+
### Training hyperparameters
|
| 47 |
+
|
| 48 |
+
The following hyperparameters were used during training:
|
| 49 |
+
- learning_rate: 5e-06
|
| 50 |
+
- train_batch_size: 4
|
| 51 |
+
- eval_batch_size: 4
|
| 52 |
+
- seed: 42
|
| 53 |
+
- distributed_type: multi-GPU
|
| 54 |
+
- num_devices: 3
|
| 55 |
+
- gradient_accumulation_steps: 12
|
| 56 |
+
- total_train_batch_size: 144
|
| 57 |
+
- total_eval_batch_size: 12
|
| 58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 59 |
+
- lr_scheduler_type: cosine
|
| 60 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 61 |
+
- num_epochs: 5
|
| 62 |
+
|
| 63 |
+
### Training results
|
| 64 |
+
|
| 65 |
+
| Training Loss | Epoch | Step | Validation Loss | Logps | Logits | Objective | Dpo Loss | Regularize | Ranking Simple | Ranking Idealized | Ranking Idealized Expo | Wo Beta |
|
| 66 |
+
|:-------------:|:------:|:----:|:---------------:|:--------:|:-------:|:---------:|:--------:|:----------:|:--------------:|:-----------------:|:----------------------:|:-------:|
|
| 67 |
+
| 2.0094 | 0.1417 | 50 | 3.1068 | -90.6242 | -1.4592 | 3.0980 | 3.0980 | 3.0980 | 0.5259 | 0.5212 | 0.5212 | 7.7179 |
|
| 68 |
+
| 5.9165 | 0.2834 | 100 | 7.1487 | -82.8335 | -1.4642 | 7.1399 | 7.1399 | 7.1399 | 0.5300 | 0.5212 | 0.5212 | 7.4498 |
|
| 69 |
+
| 9.9617 | 0.4251 | 150 | 11.8998 | -83.0745 | -1.3437 | 11.3536 | 11.3536 | 11.3536 | 0.5305 | 0.5212 | 0.5212 | 7.2609 |
|
| 70 |
+
| 12.4724 | 0.5668 | 200 | 17.0987 | -79.9360 | -1.3880 | 16.0617 | 16.0617 | 16.0617 | 0.5300 | 0.5212 | 0.5212 | 7.2290 |
|
| 71 |
+
| 13.2936 | 0.7085 | 250 | 18.5309 | -77.3150 | -1.3641 | 17.7971 | 17.7971 | 17.7971 | 0.5342 | 0.5212 | 0.5212 | 7.2078 |
|
| 72 |
+
| 11.5204 | 0.8503 | 300 | 19.4344 | -76.9798 | -0.9941 | 18.7017 | 18.7017 | 18.7017 | 0.5357 | 0.5212 | 0.5212 | 7.0136 |
|
| 73 |
+
| 11.3717 | 0.9920 | 350 | 20.3959 | -76.1623 | -1.0426 | 19.0398 | 19.0398 | 19.0398 | 0.5409 | 0.5212 | 0.5212 | 7.0261 |
|
| 74 |
+
| 7.0971 | 1.1337 | 400 | 21.9279 | -76.1458 | -0.6236 | 21.6902 | 21.6902 | 21.6902 | 0.5388 | 0.5212 | 0.5212 | 7.1227 |
|
| 75 |
+
| 7.5725 | 1.2754 | 450 | 20.9480 | -76.3924 | -0.8352 | 20.3853 | 20.3853 | 20.3853 | 0.5373 | 0.5212 | 0.5212 | 6.8500 |
|
| 76 |
+
| 7.6466 | 1.4171 | 500 | 20.9821 | -80.7806 | -0.7483 | 20.2651 | 20.2651 | 20.2651 | 0.5326 | 0.5212 | 0.5212 | 6.8824 |
|
| 77 |
+
| 6.9565 | 1.5588 | 550 | 21.3506 | -80.2051 | -0.6148 | 20.5661 | 20.5661 | 20.5661 | 0.5383 | 0.5212 | 0.5212 | 6.6513 |
|
| 78 |
+
| 6.7183 | 1.7005 | 600 | 21.1265 | -78.5344 | -0.6067 | 20.0027 | 20.0027 | 20.0027 | 0.5367 | 0.5212 | 0.5212 | 6.6768 |
|
| 79 |
+
| 6.9931 | 1.8422 | 650 | 22.2083 | -77.6509 | -0.5872 | 21.4455 | 21.4455 | 21.4455 | 0.5383 | 0.5212 | 0.5212 | 6.8190 |
|
| 80 |
+
| 6.1685 | 1.9839 | 700 | 22.3607 | -77.1493 | -0.5436 | 21.5512 | 21.5512 | 21.5512 | 0.5404 | 0.5212 | 0.5212 | 6.7299 |
|
| 81 |
+
| 3.4811 | 2.1256 | 750 | 21.8349 | -78.9312 | -0.7313 | 21.1379 | 21.1379 | 21.1379 | 0.5424 | 0.5212 | 0.5212 | 6.8213 |
|
| 82 |
+
| 3.3995 | 2.2674 | 800 | 21.3539 | -79.7115 | -0.5475 | 20.4532 | 20.4532 | 20.4532 | 0.5362 | 0.5212 | 0.5212 | 6.6867 |
|
| 83 |
+
|
| 84 |
+
|
| 85 |
+
### Framework versions
|
| 86 |
+
|
| 87 |
+
- Transformers 4.42.0
|
| 88 |
+
- Pytorch 2.3.0+cu121
|
| 89 |
+
- Datasets 2.19.1
|
| 90 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.2673594709494567,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 7.659533626437187,
|
| 5 |
+
"train_runtime": 21973.9049,
|
| 6 |
+
"train_samples": 50802,
|
| 7 |
+
"train_samples_per_second": 11.56,
|
| 8 |
+
"train_steps_per_second": 0.08
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151644,
|
| 3 |
+
"eos_token_id": 151645,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"pad_token_id": 151645,
|
| 6 |
+
"transformers_version": "4.42.0"
|
| 7 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1975192208
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9de1680114e84356068a7b1e6938567a46728c700a4677ea6c5d0c0258a70537
|
| 3 |
size 1975192208
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.2673594709494567,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 7.659533626437187,
|
| 5 |
+
"train_runtime": 21973.9049,
|
| 6 |
+
"train_samples": 50802,
|
| 7 |
+
"train_samples_per_second": 11.56,
|
| 8 |
+
"train_steps_per_second": 0.08
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,595 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 6.651296138763428,
|
| 3 |
+
"best_model_checkpoint": "./qwen2.5-0.5b/qwen2.5-0.5b-expo-DPO-ES-10/checkpoint-550",
|
| 4 |
+
"epoch": 2.2673594709494567,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"dpo_loss": 0.6931471824645996,
|
| 13 |
+
"epoch": 0.002834199338686821,
|
| 14 |
+
"grad_norm": 1844.2532039401294,
|
| 15 |
+
"learning_rate": 2.840909090909091e-08,
|
| 16 |
+
"logits": -1.359458565711975,
|
| 17 |
+
"logps": -84.69721221923828,
|
| 18 |
+
"loss": 0.6931,
|
| 19 |
+
"objective": 0.6931471824645996,
|
| 20 |
+
"ranking_idealized": 0.5833333134651184,
|
| 21 |
+
"ranking_idealized_expo": 0.5833333134651184,
|
| 22 |
+
"ranking_simple": 0.5833333134651184,
|
| 23 |
+
"regularize": 0.6931471824645996,
|
| 24 |
+
"step": 1,
|
| 25 |
+
"wo_beta": 5.271125316619873
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"dpo_loss": 1.9794068336486816,
|
| 29 |
+
"epoch": 0.14170996693434104,
|
| 30 |
+
"grad_norm": 1879.680280823908,
|
| 31 |
+
"learning_rate": 1.4204545454545458e-06,
|
| 32 |
+
"logits": -1.4480701684951782,
|
| 33 |
+
"logps": -84.59326934814453,
|
| 34 |
+
"loss": 2.0094,
|
| 35 |
+
"objective": 1.9794068336486816,
|
| 36 |
+
"ranking_idealized": 0.5225340127944946,
|
| 37 |
+
"ranking_idealized_expo": 0.5216836929321289,
|
| 38 |
+
"ranking_simple": 0.5250850319862366,
|
| 39 |
+
"regularize": 1.9794068336486816,
|
| 40 |
+
"step": 50,
|
| 41 |
+
"wo_beta": 7.08821439743042
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.14170996693434104,
|
| 45 |
+
"eval_dpo_loss": 3.0980334281921387,
|
| 46 |
+
"eval_logits": -1.4591896533966064,
|
| 47 |
+
"eval_logps": -90.62417602539062,
|
| 48 |
+
"eval_loss": 3.106841564178467,
|
| 49 |
+
"eval_objective": 3.0980334281921387,
|
| 50 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 51 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 52 |
+
"eval_ranking_simple": 0.5258799195289612,
|
| 53 |
+
"eval_regularize": 3.0980334281921387,
|
| 54 |
+
"eval_runtime": 308.1993,
|
| 55 |
+
"eval_samples_per_second": 18.787,
|
| 56 |
+
"eval_steps_per_second": 1.567,
|
| 57 |
+
"eval_wo_beta": 7.7179274559021,
|
| 58 |
+
"step": 50
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"dpo_loss": 6.002392768859863,
|
| 62 |
+
"epoch": 0.2834199338686821,
|
| 63 |
+
"grad_norm": 1776.9003571892035,
|
| 64 |
+
"learning_rate": 2.8409090909090916e-06,
|
| 65 |
+
"logits": -1.4501550197601318,
|
| 66 |
+
"logps": -81.94203186035156,
|
| 67 |
+
"loss": 5.9165,
|
| 68 |
+
"objective": 6.002392768859863,
|
| 69 |
+
"ranking_idealized": 0.5141666531562805,
|
| 70 |
+
"ranking_idealized_expo": 0.5137500166893005,
|
| 71 |
+
"ranking_simple": 0.5425000190734863,
|
| 72 |
+
"regularize": 6.002392768859863,
|
| 73 |
+
"step": 100,
|
| 74 |
+
"wo_beta": 6.498049736022949
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.2834199338686821,
|
| 78 |
+
"eval_dpo_loss": 7.139862060546875,
|
| 79 |
+
"eval_logits": -1.4642183780670166,
|
| 80 |
+
"eval_logps": -82.83346557617188,
|
| 81 |
+
"eval_loss": 7.14874792098999,
|
| 82 |
+
"eval_objective": 7.139862060546875,
|
| 83 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 84 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 85 |
+
"eval_ranking_simple": 0.5300207138061523,
|
| 86 |
+
"eval_regularize": 7.139862060546875,
|
| 87 |
+
"eval_runtime": 309.8222,
|
| 88 |
+
"eval_samples_per_second": 18.688,
|
| 89 |
+
"eval_steps_per_second": 1.559,
|
| 90 |
+
"eval_wo_beta": 7.4498443603515625,
|
| 91 |
+
"step": 100
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"dpo_loss": 10.150534629821777,
|
| 95 |
+
"epoch": 0.42512990080302315,
|
| 96 |
+
"grad_norm": 1372.4824531102197,
|
| 97 |
+
"learning_rate": 4.2613636363636365e-06,
|
| 98 |
+
"logits": -1.4184441566467285,
|
| 99 |
+
"logps": -73.7444076538086,
|
| 100 |
+
"loss": 9.9617,
|
| 101 |
+
"objective": 10.150534629821777,
|
| 102 |
+
"ranking_idealized": 0.5287500023841858,
|
| 103 |
+
"ranking_idealized_expo": 0.527916669845581,
|
| 104 |
+
"ranking_simple": 0.5641666650772095,
|
| 105 |
+
"regularize": 10.150534629821777,
|
| 106 |
+
"step": 150,
|
| 107 |
+
"wo_beta": 6.36607027053833
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.42512990080302315,
|
| 111 |
+
"eval_dpo_loss": 11.353630065917969,
|
| 112 |
+
"eval_logits": -1.3437175750732422,
|
| 113 |
+
"eval_logps": -83.07452392578125,
|
| 114 |
+
"eval_loss": 11.899770736694336,
|
| 115 |
+
"eval_objective": 11.353630065917969,
|
| 116 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 117 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 118 |
+
"eval_ranking_simple": 0.5305383205413818,
|
| 119 |
+
"eval_regularize": 11.353630065917969,
|
| 120 |
+
"eval_runtime": 307.9181,
|
| 121 |
+
"eval_samples_per_second": 18.804,
|
| 122 |
+
"eval_steps_per_second": 1.569,
|
| 123 |
+
"eval_wo_beta": 7.260918140411377,
|
| 124 |
+
"step": 150
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"dpo_loss": 12.522791862487793,
|
| 128 |
+
"epoch": 0.5668398677373642,
|
| 129 |
+
"grad_norm": 1378.4122138720427,
|
| 130 |
+
"learning_rate": 4.997168347957521e-06,
|
| 131 |
+
"logits": -1.3766180276870728,
|
| 132 |
+
"logps": -77.5620346069336,
|
| 133 |
+
"loss": 12.4724,
|
| 134 |
+
"objective": 12.522791862487793,
|
| 135 |
+
"ranking_idealized": 0.51583331823349,
|
| 136 |
+
"ranking_idealized_expo": 0.51541668176651,
|
| 137 |
+
"ranking_simple": 0.5550000071525574,
|
| 138 |
+
"regularize": 12.522791862487793,
|
| 139 |
+
"step": 200,
|
| 140 |
+
"wo_beta": 6.352013111114502
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.5668398677373642,
|
| 144 |
+
"eval_dpo_loss": 16.061721801757812,
|
| 145 |
+
"eval_logits": -1.387966513633728,
|
| 146 |
+
"eval_logps": -79.93595886230469,
|
| 147 |
+
"eval_loss": 17.09868049621582,
|
| 148 |
+
"eval_objective": 16.061721801757812,
|
| 149 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 150 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 151 |
+
"eval_ranking_simple": 0.5300207138061523,
|
| 152 |
+
"eval_regularize": 16.061721801757812,
|
| 153 |
+
"eval_runtime": 307.8039,
|
| 154 |
+
"eval_samples_per_second": 18.811,
|
| 155 |
+
"eval_steps_per_second": 1.569,
|
| 156 |
+
"eval_wo_beta": 7.228997230529785,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"dpo_loss": 13.093570709228516,
|
| 161 |
+
"epoch": 0.7085498346717053,
|
| 162 |
+
"grad_norm": 1274.583157442186,
|
| 163 |
+
"learning_rate": 4.973122855144066e-06,
|
| 164 |
+
"logits": -1.294631004333496,
|
| 165 |
+
"logps": -76.13822174072266,
|
| 166 |
+
"loss": 13.2936,
|
| 167 |
+
"objective": 13.093570709228516,
|
| 168 |
+
"ranking_idealized": 0.5166666507720947,
|
| 169 |
+
"ranking_idealized_expo": 0.5162500143051147,
|
| 170 |
+
"ranking_simple": 0.5824999809265137,
|
| 171 |
+
"regularize": 13.093570709228516,
|
| 172 |
+
"step": 250,
|
| 173 |
+
"wo_beta": 6.279551982879639
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.7085498346717053,
|
| 177 |
+
"eval_dpo_loss": 17.797138214111328,
|
| 178 |
+
"eval_logits": -1.3640648126602173,
|
| 179 |
+
"eval_logps": -77.31498718261719,
|
| 180 |
+
"eval_loss": 18.5308780670166,
|
| 181 |
+
"eval_objective": 17.797138214111328,
|
| 182 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 183 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 184 |
+
"eval_ranking_simple": 0.5341615080833435,
|
| 185 |
+
"eval_regularize": 17.797138214111328,
|
| 186 |
+
"eval_runtime": 313.1016,
|
| 187 |
+
"eval_samples_per_second": 18.492,
|
| 188 |
+
"eval_steps_per_second": 1.543,
|
| 189 |
+
"eval_wo_beta": 7.207766532897949,
|
| 190 |
+
"step": 250
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"dpo_loss": 10.712362289428711,
|
| 194 |
+
"epoch": 0.8502598016060463,
|
| 195 |
+
"grad_norm": 982.3462926804266,
|
| 196 |
+
"learning_rate": 4.924776641419513e-06,
|
| 197 |
+
"logits": -1.090299129486084,
|
| 198 |
+
"logps": -70.98873138427734,
|
| 199 |
+
"loss": 11.5204,
|
| 200 |
+
"objective": 10.712362289428711,
|
| 201 |
+
"ranking_idealized": 0.4962500035762787,
|
| 202 |
+
"ranking_idealized_expo": 0.4950000047683716,
|
| 203 |
+
"ranking_simple": 0.5679166913032532,
|
| 204 |
+
"regularize": 10.712362289428711,
|
| 205 |
+
"step": 300,
|
| 206 |
+
"wo_beta": 6.134185314178467
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.8502598016060463,
|
| 210 |
+
"eval_dpo_loss": 18.701662063598633,
|
| 211 |
+
"eval_logits": -0.9941285848617554,
|
| 212 |
+
"eval_logps": -76.97978210449219,
|
| 213 |
+
"eval_loss": 19.434432983398438,
|
| 214 |
+
"eval_objective": 18.701662063598633,
|
| 215 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 216 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 217 |
+
"eval_ranking_simple": 0.5357142686843872,
|
| 218 |
+
"eval_regularize": 18.701662063598633,
|
| 219 |
+
"eval_runtime": 307.6602,
|
| 220 |
+
"eval_samples_per_second": 18.819,
|
| 221 |
+
"eval_steps_per_second": 1.57,
|
| 222 |
+
"eval_wo_beta": 7.013552188873291,
|
| 223 |
+
"step": 300
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"dpo_loss": 10.92597484588623,
|
| 227 |
+
"epoch": 0.9919697685403873,
|
| 228 |
+
"grad_norm": 938.8397527375307,
|
| 229 |
+
"learning_rate": 4.8526047530778175e-06,
|
| 230 |
+
"logits": -0.9006206393241882,
|
| 231 |
+
"logps": -72.82616424560547,
|
| 232 |
+
"loss": 11.3717,
|
| 233 |
+
"objective": 10.92597484588623,
|
| 234 |
+
"ranking_idealized": 0.5262500047683716,
|
| 235 |
+
"ranking_idealized_expo": 0.5254166722297668,
|
| 236 |
+
"ranking_simple": 0.6033333539962769,
|
| 237 |
+
"regularize": 10.92597484588623,
|
| 238 |
+
"step": 350,
|
| 239 |
+
"wo_beta": 5.362515449523926
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.9919697685403873,
|
| 243 |
+
"eval_dpo_loss": 19.039833068847656,
|
| 244 |
+
"eval_logits": -1.0426429510116577,
|
| 245 |
+
"eval_logps": -76.1622543334961,
|
| 246 |
+
"eval_loss": 20.39594841003418,
|
| 247 |
+
"eval_objective": 19.039833068847656,
|
| 248 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 249 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 250 |
+
"eval_ranking_simple": 0.5408902764320374,
|
| 251 |
+
"eval_regularize": 19.039833068847656,
|
| 252 |
+
"eval_runtime": 307.6502,
|
| 253 |
+
"eval_samples_per_second": 18.82,
|
| 254 |
+
"eval_steps_per_second": 1.57,
|
| 255 |
+
"eval_wo_beta": 7.0260910987854,
|
| 256 |
+
"step": 350
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"dpo_loss": 6.708657264709473,
|
| 260 |
+
"epoch": 1.1336797354747283,
|
| 261 |
+
"grad_norm": 902.5480798954853,
|
| 262 |
+
"learning_rate": 4.757316345716554e-06,
|
| 263 |
+
"logits": -0.7562137246131897,
|
| 264 |
+
"logps": -70.6362075805664,
|
| 265 |
+
"loss": 7.0971,
|
| 266 |
+
"objective": 6.708657264709473,
|
| 267 |
+
"ranking_idealized": 0.5333333611488342,
|
| 268 |
+
"ranking_idealized_expo": 0.5320833325386047,
|
| 269 |
+
"ranking_simple": 0.6329166889190674,
|
| 270 |
+
"regularize": 6.708657264709473,
|
| 271 |
+
"step": 400,
|
| 272 |
+
"wo_beta": 4.74643087387085
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.1336797354747283,
|
| 276 |
+
"eval_dpo_loss": 21.69021987915039,
|
| 277 |
+
"eval_logits": -0.6236207485198975,
|
| 278 |
+
"eval_logps": -76.14582824707031,
|
| 279 |
+
"eval_loss": 21.927854537963867,
|
| 280 |
+
"eval_objective": 21.69021987915039,
|
| 281 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 282 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 283 |
+
"eval_ranking_simple": 0.5388198494911194,
|
| 284 |
+
"eval_regularize": 21.69021987915039,
|
| 285 |
+
"eval_runtime": 308.4451,
|
| 286 |
+
"eval_samples_per_second": 18.772,
|
| 287 |
+
"eval_steps_per_second": 1.566,
|
| 288 |
+
"eval_wo_beta": 7.122740745544434,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"dpo_loss": 8.22778606414795,
|
| 293 |
+
"epoch": 1.2753897024090695,
|
| 294 |
+
"grad_norm": 701.1530681925066,
|
| 295 |
+
"learning_rate": 4.639847716126855e-06,
|
| 296 |
+
"logits": -0.6124467849731445,
|
| 297 |
+
"logps": -71.35508728027344,
|
| 298 |
+
"loss": 7.5725,
|
| 299 |
+
"objective": 8.22778606414795,
|
| 300 |
+
"ranking_idealized": 0.5195833444595337,
|
| 301 |
+
"ranking_idealized_expo": 0.5191666483879089,
|
| 302 |
+
"ranking_simple": 0.6312500238418579,
|
| 303 |
+
"regularize": 8.22778606414795,
|
| 304 |
+
"step": 450,
|
| 305 |
+
"wo_beta": 5.267808437347412
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.2753897024090695,
|
| 309 |
+
"eval_dpo_loss": 20.385303497314453,
|
| 310 |
+
"eval_logits": -0.8352137207984924,
|
| 311 |
+
"eval_logps": -76.3924331665039,
|
| 312 |
+
"eval_loss": 20.948013305664062,
|
| 313 |
+
"eval_objective": 20.385303497314453,
|
| 314 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 315 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 316 |
+
"eval_ranking_simple": 0.5372670888900757,
|
| 317 |
+
"eval_regularize": 20.385303497314453,
|
| 318 |
+
"eval_runtime": 307.7791,
|
| 319 |
+
"eval_samples_per_second": 18.812,
|
| 320 |
+
"eval_steps_per_second": 1.569,
|
| 321 |
+
"eval_wo_beta": 6.8499908447265625,
|
| 322 |
+
"step": 450
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"dpo_loss": 7.319465160369873,
|
| 326 |
+
"epoch": 1.4170996693434104,
|
| 327 |
+
"grad_norm": 1027.4391137177338,
|
| 328 |
+
"learning_rate": 4.501353102310901e-06,
|
| 329 |
+
"logits": -0.7022644877433777,
|
| 330 |
+
"logps": -74.45861053466797,
|
| 331 |
+
"loss": 7.6466,
|
| 332 |
+
"objective": 7.319465160369873,
|
| 333 |
+
"ranking_idealized": 0.49791666865348816,
|
| 334 |
+
"ranking_idealized_expo": 0.4970833361148834,
|
| 335 |
+
"ranking_simple": 0.6225000023841858,
|
| 336 |
+
"regularize": 7.319465160369873,
|
| 337 |
+
"step": 500,
|
| 338 |
+
"wo_beta": 5.078485488891602
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4170996693434104,
|
| 342 |
+
"eval_dpo_loss": 20.265100479125977,
|
| 343 |
+
"eval_logits": -0.7483307123184204,
|
| 344 |
+
"eval_logps": -80.78058624267578,
|
| 345 |
+
"eval_loss": 20.982105255126953,
|
| 346 |
+
"eval_objective": 20.265100479125977,
|
| 347 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 348 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 349 |
+
"eval_ranking_simple": 0.532608687877655,
|
| 350 |
+
"eval_regularize": 20.265100479125977,
|
| 351 |
+
"eval_runtime": 307.6224,
|
| 352 |
+
"eval_samples_per_second": 18.822,
|
| 353 |
+
"eval_steps_per_second": 1.57,
|
| 354 |
+
"eval_wo_beta": 6.882425785064697,
|
| 355 |
+
"step": 500
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"dpo_loss": 6.880460739135742,
|
| 359 |
+
"epoch": 1.5588096362777515,
|
| 360 |
+
"grad_norm": 856.3100755197052,
|
| 361 |
+
"learning_rate": 4.34319334202531e-06,
|
| 362 |
+
"logits": -0.6065574884414673,
|
| 363 |
+
"logps": -75.99507141113281,
|
| 364 |
+
"loss": 6.9565,
|
| 365 |
+
"objective": 6.880460739135742,
|
| 366 |
+
"ranking_idealized": 0.5112500190734863,
|
| 367 |
+
"ranking_idealized_expo": 0.5104166865348816,
|
| 368 |
+
"ranking_simple": 0.6445833444595337,
|
| 369 |
+
"regularize": 6.880460739135742,
|
| 370 |
+
"step": 550,
|
| 371 |
+
"wo_beta": 4.776731967926025
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.5588096362777515,
|
| 375 |
+
"eval_dpo_loss": 20.566144943237305,
|
| 376 |
+
"eval_logits": -0.6148493885993958,
|
| 377 |
+
"eval_logps": -80.20514678955078,
|
| 378 |
+
"eval_loss": 21.350601196289062,
|
| 379 |
+
"eval_objective": 20.566144943237305,
|
| 380 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 381 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 382 |
+
"eval_ranking_simple": 0.5383023023605347,
|
| 383 |
+
"eval_regularize": 20.566144943237305,
|
| 384 |
+
"eval_runtime": 311.6281,
|
| 385 |
+
"eval_samples_per_second": 18.58,
|
| 386 |
+
"eval_steps_per_second": 1.55,
|
| 387 |
+
"eval_wo_beta": 6.651296138763428,
|
| 388 |
+
"step": 550
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"dpo_loss": 6.715544700622559,
|
| 392 |
+
"epoch": 1.7005196032120926,
|
| 393 |
+
"grad_norm": 710.120603889053,
|
| 394 |
+
"learning_rate": 4.16692250129073e-06,
|
| 395 |
+
"logits": -0.4755525290966034,
|
| 396 |
+
"logps": -75.72002410888672,
|
| 397 |
+
"loss": 6.7183,
|
| 398 |
+
"objective": 6.715544700622559,
|
| 399 |
+
"ranking_idealized": 0.51541668176651,
|
| 400 |
+
"ranking_idealized_expo": 0.5149999856948853,
|
| 401 |
+
"ranking_simple": 0.637499988079071,
|
| 402 |
+
"regularize": 6.715544700622559,
|
| 403 |
+
"step": 600,
|
| 404 |
+
"wo_beta": 4.843540668487549
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.7005196032120926,
|
| 408 |
+
"eval_dpo_loss": 20.002656936645508,
|
| 409 |
+
"eval_logits": -0.606741189956665,
|
| 410 |
+
"eval_logps": -78.53438568115234,
|
| 411 |
+
"eval_loss": 21.126510620117188,
|
| 412 |
+
"eval_objective": 20.002656936645508,
|
| 413 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 414 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 415 |
+
"eval_ranking_simple": 0.5367494821548462,
|
| 416 |
+
"eval_regularize": 20.002656936645508,
|
| 417 |
+
"eval_runtime": 307.6088,
|
| 418 |
+
"eval_samples_per_second": 18.823,
|
| 419 |
+
"eval_steps_per_second": 1.57,
|
| 420 |
+
"eval_wo_beta": 6.676760673522949,
|
| 421 |
+
"step": 600
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"dpo_loss": 7.323308944702148,
|
| 425 |
+
"epoch": 1.8422295701464337,
|
| 426 |
+
"grad_norm": 790.4876193704064,
|
| 427 |
+
"learning_rate": 3.974272604254906e-06,
|
| 428 |
+
"logits": -0.5003318190574646,
|
| 429 |
+
"logps": -74.34846496582031,
|
| 430 |
+
"loss": 6.9931,
|
| 431 |
+
"objective": 7.323308944702148,
|
| 432 |
+
"ranking_idealized": 0.5291666388511658,
|
| 433 |
+
"ranking_idealized_expo": 0.527916669845581,
|
| 434 |
+
"ranking_simple": 0.64083331823349,
|
| 435 |
+
"regularize": 7.323308944702148,
|
| 436 |
+
"step": 650,
|
| 437 |
+
"wo_beta": 5.1112799644470215
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.8422295701464337,
|
| 441 |
+
"eval_dpo_loss": 21.445514678955078,
|
| 442 |
+
"eval_logits": -0.5872498154640198,
|
| 443 |
+
"eval_logps": -77.65087127685547,
|
| 444 |
+
"eval_loss": 22.20830535888672,
|
| 445 |
+
"eval_objective": 21.445514678955078,
|
| 446 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 447 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 448 |
+
"eval_ranking_simple": 0.5383023023605347,
|
| 449 |
+
"eval_regularize": 21.445514678955078,
|
| 450 |
+
"eval_runtime": 307.7733,
|
| 451 |
+
"eval_samples_per_second": 18.813,
|
| 452 |
+
"eval_steps_per_second": 1.569,
|
| 453 |
+
"eval_wo_beta": 6.819047451019287,
|
| 454 |
+
"step": 650
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"dpo_loss": 6.620248317718506,
|
| 458 |
+
"epoch": 1.9839395370807746,
|
| 459 |
+
"grad_norm": 688.8652687295252,
|
| 460 |
+
"learning_rate": 3.767136614452458e-06,
|
| 461 |
+
"logits": -0.40135031938552856,
|
| 462 |
+
"logps": -73.09497833251953,
|
| 463 |
+
"loss": 6.1685,
|
| 464 |
+
"objective": 6.620248317718506,
|
| 465 |
+
"ranking_idealized": 0.5129166841506958,
|
| 466 |
+
"ranking_idealized_expo": 0.5108333230018616,
|
| 467 |
+
"ranking_simple": 0.6358333230018616,
|
| 468 |
+
"regularize": 6.620248317718506,
|
| 469 |
+
"step": 700,
|
| 470 |
+
"wo_beta": 5.023129463195801
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.9839395370807746,
|
| 474 |
+
"eval_dpo_loss": 21.55119514465332,
|
| 475 |
+
"eval_logits": -0.5436362028121948,
|
| 476 |
+
"eval_logps": -77.14934539794922,
|
| 477 |
+
"eval_loss": 22.36069679260254,
|
| 478 |
+
"eval_objective": 21.55119514465332,
|
| 479 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 480 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 481 |
+
"eval_ranking_simple": 0.5403726696968079,
|
| 482 |
+
"eval_regularize": 21.55119514465332,
|
| 483 |
+
"eval_runtime": 310.6834,
|
| 484 |
+
"eval_samples_per_second": 18.636,
|
| 485 |
+
"eval_steps_per_second": 1.555,
|
| 486 |
+
"eval_wo_beta": 6.729911804199219,
|
| 487 |
+
"step": 700
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"dpo_loss": 3.5603878498077393,
|
| 491 |
+
"epoch": 2.1256495040151155,
|
| 492 |
+
"grad_norm": 579.769858214478,
|
| 493 |
+
"learning_rate": 3.547549834686222e-06,
|
| 494 |
+
"logits": -0.5370141863822937,
|
| 495 |
+
"logps": -73.9045639038086,
|
| 496 |
+
"loss": 3.4811,
|
| 497 |
+
"objective": 3.5603878498077393,
|
| 498 |
+
"ranking_idealized": 0.5129166841506958,
|
| 499 |
+
"ranking_idealized_expo": 0.5112500190734863,
|
| 500 |
+
"ranking_simple": 0.6691666841506958,
|
| 501 |
+
"regularize": 3.5603878498077393,
|
| 502 |
+
"step": 750,
|
| 503 |
+
"wo_beta": 4.534417152404785
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.1256495040151155,
|
| 507 |
+
"eval_dpo_loss": 21.137874603271484,
|
| 508 |
+
"eval_logits": -0.7312601804733276,
|
| 509 |
+
"eval_logps": -78.93118286132812,
|
| 510 |
+
"eval_loss": 21.834890365600586,
|
| 511 |
+
"eval_objective": 21.137874603271484,
|
| 512 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 513 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 514 |
+
"eval_ranking_simple": 0.542443037033081,
|
| 515 |
+
"eval_regularize": 21.137874603271484,
|
| 516 |
+
"eval_runtime": 307.6064,
|
| 517 |
+
"eval_samples_per_second": 18.823,
|
| 518 |
+
"eval_steps_per_second": 1.57,
|
| 519 |
+
"eval_wo_beta": 6.821295261383057,
|
| 520 |
+
"step": 750
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"dpo_loss": 3.4036636352539062,
|
| 524 |
+
"epoch": 2.2673594709494567,
|
| 525 |
+
"grad_norm": 573.1633009551587,
|
| 526 |
+
"learning_rate": 3.3176699082935546e-06,
|
| 527 |
+
"logits": -0.5852146148681641,
|
| 528 |
+
"logps": -75.8536376953125,
|
| 529 |
+
"loss": 3.3995,
|
| 530 |
+
"objective": 3.4036636352539062,
|
| 531 |
+
"ranking_idealized": 0.512499988079071,
|
| 532 |
+
"ranking_idealized_expo": 0.512499988079071,
|
| 533 |
+
"ranking_simple": 0.6625000238418579,
|
| 534 |
+
"regularize": 3.4036636352539062,
|
| 535 |
+
"step": 800,
|
| 536 |
+
"wo_beta": 4.245257377624512
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.2673594709494567,
|
| 540 |
+
"eval_dpo_loss": 20.453168869018555,
|
| 541 |
+
"eval_logits": -0.5475257635116577,
|
| 542 |
+
"eval_logps": -79.71145629882812,
|
| 543 |
+
"eval_loss": 21.353944778442383,
|
| 544 |
+
"eval_objective": 20.453168869018555,
|
| 545 |
+
"eval_ranking_idealized": 0.5212215185165405,
|
| 546 |
+
"eval_ranking_idealized_expo": 0.5212215185165405,
|
| 547 |
+
"eval_ranking_simple": 0.5362318754196167,
|
| 548 |
+
"eval_regularize": 20.453168869018555,
|
| 549 |
+
"eval_runtime": 307.5814,
|
| 550 |
+
"eval_samples_per_second": 18.824,
|
| 551 |
+
"eval_steps_per_second": 1.57,
|
| 552 |
+
"eval_wo_beta": 6.686735153198242,
|
| 553 |
+
"step": 800
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.2673594709494567,
|
| 557 |
+
"step": 800,
|
| 558 |
+
"total_flos": 0.0,
|
| 559 |
+
"train_loss": 7.659533626437187,
|
| 560 |
+
"train_runtime": 21973.9049,
|
| 561 |
+
"train_samples_per_second": 11.56,
|
| 562 |
+
"train_steps_per_second": 0.08
|
| 563 |
+
}
|
| 564 |
+
],
|
| 565 |
+
"logging_steps": 50,
|
| 566 |
+
"max_steps": 1760,
|
| 567 |
+
"num_input_tokens_seen": 0,
|
| 568 |
+
"num_train_epochs": 5,
|
| 569 |
+
"save_steps": 50,
|
| 570 |
+
"stateful_callbacks": {
|
| 571 |
+
"EarlyStoppingCallback": {
|
| 572 |
+
"args": {
|
| 573 |
+
"early_stopping_patience": 5,
|
| 574 |
+
"early_stopping_threshold": 0.0
|
| 575 |
+
},
|
| 576 |
+
"attributes": {
|
| 577 |
+
"early_stopping_patience_counter": 0
|
| 578 |
+
}
|
| 579 |
+
},
|
| 580 |
+
"TrainerControl": {
|
| 581 |
+
"args": {
|
| 582 |
+
"should_epoch_stop": false,
|
| 583 |
+
"should_evaluate": false,
|
| 584 |
+
"should_log": false,
|
| 585 |
+
"should_save": true,
|
| 586 |
+
"should_training_stop": true
|
| 587 |
+
},
|
| 588 |
+
"attributes": {}
|
| 589 |
+
}
|
| 590 |
+
},
|
| 591 |
+
"total_flos": 0.0,
|
| 592 |
+
"train_batch_size": 4,
|
| 593 |
+
"trial_name": null,
|
| 594 |
+
"trial_params": null
|
| 595 |
+
}
|