Model save
Browse files- README.md +60 -0
- all_results.json +9 -0
- generation_config.json +12 -0
- train_results.json +9 -0
- trainer_state.json +987 -0
README.md
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: llama3
|
| 3 |
+
base_model: Jackie999/llama3-sudo-5epochs-tofu_full_sft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
model-index:
|
| 9 |
+
- name: llama3-sudo-dpo-5epochs-forget10mixall-1sft-2fullpara-1e-5
|
| 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 |
+
# llama3-sudo-dpo-5epochs-forget10mixall-1sft-2fullpara-1e-5
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [Jackie999/llama3-sudo-5epochs-tofu_full_sft](https://huggingface.co/Jackie999/llama3-sudo-5epochs-tofu_full_sft) on an unknown dataset.
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 1e-05
|
| 38 |
+
- train_batch_size: 4
|
| 39 |
+
- eval_batch_size: 8
|
| 40 |
+
- seed: 42
|
| 41 |
+
- distributed_type: multi-GPU
|
| 42 |
+
- num_devices: 4
|
| 43 |
+
- gradient_accumulation_steps: 4
|
| 44 |
+
- total_train_batch_size: 64
|
| 45 |
+
- total_eval_batch_size: 32
|
| 46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
+
- lr_scheduler_type: cosine
|
| 48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 49 |
+
- num_epochs: 5
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
### Framework versions
|
| 56 |
+
|
| 57 |
+
- Transformers 4.44.0
|
| 58 |
+
- Pytorch 2.1.2
|
| 59 |
+
- Datasets 3.0.0
|
| 60 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.022511796173290348,
|
| 5 |
+
"train_runtime": 65461.2558,
|
| 6 |
+
"train_samples": 8000,
|
| 7 |
+
"train_samples_per_second": 0.611,
|
| 8 |
+
"train_steps_per_second": 0.01
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 128000,
|
| 3 |
+
"do_sample": true,
|
| 4 |
+
"eos_token_id": [
|
| 5 |
+
128001,
|
| 6 |
+
128009
|
| 7 |
+
],
|
| 8 |
+
"max_length": 4096,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "4.44.0"
|
| 12 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.022511796173290348,
|
| 5 |
+
"train_runtime": 65461.2558,
|
| 6 |
+
"train_samples": 8000,
|
| 7 |
+
"train_samples_per_second": 0.611,
|
| 8 |
+
"train_steps_per_second": 0.01
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,987 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 625,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008,
|
| 13 |
+
"grad_norm": 7.598883452709097,
|
| 14 |
+
"learning_rate": 1.5873015873015874e-07,
|
| 15 |
+
"logits/chosen": -1.7671998739242554,
|
| 16 |
+
"logits/rejected": -2.2639822959899902,
|
| 17 |
+
"logps/chosen": -46.430763244628906,
|
| 18 |
+
"logps/rejected": -102.85381317138672,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08,
|
| 28 |
+
"grad_norm": 9.021711112045313,
|
| 29 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 30 |
+
"logits/chosen": -1.6467827558517456,
|
| 31 |
+
"logits/rejected": -2.05173921585083,
|
| 32 |
+
"logps/chosen": -50.05971145629883,
|
| 33 |
+
"logps/rejected": -110.39069366455078,
|
| 34 |
+
"loss": 0.6403,
|
| 35 |
+
"rewards/accuracies": 0.7777777910232544,
|
| 36 |
+
"rewards/chosen": 0.0030777277424931526,
|
| 37 |
+
"rewards/margins": 0.11931005865335464,
|
| 38 |
+
"rewards/rejected": -0.11623234301805496,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.16,
|
| 43 |
+
"grad_norm": 1.8190332435996819,
|
| 44 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 45 |
+
"logits/chosen": -1.3254443407058716,
|
| 46 |
+
"logits/rejected": -1.779170036315918,
|
| 47 |
+
"logps/chosen": -37.68037033081055,
|
| 48 |
+
"logps/rejected": -313.1996154785156,
|
| 49 |
+
"loss": 0.2472,
|
| 50 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 51 |
+
"rewards/chosen": 0.09366801381111145,
|
| 52 |
+
"rewards/margins": 2.184483051300049,
|
| 53 |
+
"rewards/rejected": -2.0908150672912598,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.24,
|
| 58 |
+
"grad_norm": 8.393638030400462,
|
| 59 |
+
"learning_rate": 4.761904761904762e-06,
|
| 60 |
+
"logits/chosen": -0.7945032715797424,
|
| 61 |
+
"logits/rejected": -1.1047414541244507,
|
| 62 |
+
"logps/chosen": -55.5176887512207,
|
| 63 |
+
"logps/rejected": -406.99609375,
|
| 64 |
+
"loss": 0.1799,
|
| 65 |
+
"rewards/accuracies": 0.925000011920929,
|
| 66 |
+
"rewards/chosen": -0.04824959114193916,
|
| 67 |
+
"rewards/margins": 3.0284643173217773,
|
| 68 |
+
"rewards/rejected": -3.076713800430298,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.32,
|
| 73 |
+
"grad_norm": 3.6170805565428488,
|
| 74 |
+
"learning_rate": 6.349206349206349e-06,
|
| 75 |
+
"logits/chosen": -0.3710061311721802,
|
| 76 |
+
"logits/rejected": -0.7456581592559814,
|
| 77 |
+
"logps/chosen": -50.16427230834961,
|
| 78 |
+
"logps/rejected": -541.6275634765625,
|
| 79 |
+
"loss": 0.1374,
|
| 80 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 81 |
+
"rewards/chosen": -0.01431644894182682,
|
| 82 |
+
"rewards/margins": 4.384586334228516,
|
| 83 |
+
"rewards/rejected": -4.398902416229248,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.4,
|
| 88 |
+
"grad_norm": 3.795764554921293,
|
| 89 |
+
"learning_rate": 7.936507936507936e-06,
|
| 90 |
+
"logits/chosen": -0.581534206867218,
|
| 91 |
+
"logits/rejected": -0.8648965954780579,
|
| 92 |
+
"logps/chosen": -80.1789779663086,
|
| 93 |
+
"logps/rejected": -672.813720703125,
|
| 94 |
+
"loss": 0.0553,
|
| 95 |
+
"rewards/accuracies": 0.987500011920929,
|
| 96 |
+
"rewards/chosen": -0.3146038353443146,
|
| 97 |
+
"rewards/margins": 5.392758846282959,
|
| 98 |
+
"rewards/rejected": -5.707362174987793,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.48,
|
| 103 |
+
"grad_norm": 1.397363269705544,
|
| 104 |
+
"learning_rate": 9.523809523809525e-06,
|
| 105 |
+
"logits/chosen": -0.7689735293388367,
|
| 106 |
+
"logits/rejected": -0.9561458826065063,
|
| 107 |
+
"logps/chosen": -145.53469848632812,
|
| 108 |
+
"logps/rejected": -831.8748779296875,
|
| 109 |
+
"loss": 0.0371,
|
| 110 |
+
"rewards/accuracies": 1.0,
|
| 111 |
+
"rewards/chosen": -0.9438420534133911,
|
| 112 |
+
"rewards/margins": 6.384782314300537,
|
| 113 |
+
"rewards/rejected": -7.328624725341797,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.56,
|
| 118 |
+
"grad_norm": 0.13618082387486424,
|
| 119 |
+
"learning_rate": 9.996172565322375e-06,
|
| 120 |
+
"logits/chosen": -1.2459557056427002,
|
| 121 |
+
"logits/rejected": -1.4872663021087646,
|
| 122 |
+
"logps/chosen": -132.92080688476562,
|
| 123 |
+
"logps/rejected": -911.9078979492188,
|
| 124 |
+
"loss": 0.0409,
|
| 125 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 126 |
+
"rewards/chosen": -0.8237069249153137,
|
| 127 |
+
"rewards/margins": 7.28142786026001,
|
| 128 |
+
"rewards/rejected": -8.105135917663574,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.64,
|
| 133 |
+
"grad_norm": 1.3606359582288166,
|
| 134 |
+
"learning_rate": 9.97744005136599e-06,
|
| 135 |
+
"logits/chosen": -1.503824234008789,
|
| 136 |
+
"logits/rejected": -1.8992702960968018,
|
| 137 |
+
"logps/chosen": -115.5097885131836,
|
| 138 |
+
"logps/rejected": -916.9987182617188,
|
| 139 |
+
"loss": 0.013,
|
| 140 |
+
"rewards/accuracies": 1.0,
|
| 141 |
+
"rewards/chosen": -0.6468590497970581,
|
| 142 |
+
"rewards/margins": 7.513754367828369,
|
| 143 |
+
"rewards/rejected": -8.160614013671875,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.72,
|
| 148 |
+
"grad_norm": 8.135027509581384,
|
| 149 |
+
"learning_rate": 9.943157907471825e-06,
|
| 150 |
+
"logits/chosen": -1.5570601224899292,
|
| 151 |
+
"logits/rejected": -1.9588285684585571,
|
| 152 |
+
"logps/chosen": -122.7470703125,
|
| 153 |
+
"logps/rejected": -1007.5989379882812,
|
| 154 |
+
"loss": 0.0185,
|
| 155 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 156 |
+
"rewards/chosen": -0.71613609790802,
|
| 157 |
+
"rewards/margins": 8.348057746887207,
|
| 158 |
+
"rewards/rejected": -9.064192771911621,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.8,
|
| 163 |
+
"grad_norm": 0.07850356571109195,
|
| 164 |
+
"learning_rate": 9.893433231795864e-06,
|
| 165 |
+
"logits/chosen": -2.478653907775879,
|
| 166 |
+
"logits/rejected": -2.867173671722412,
|
| 167 |
+
"logps/chosen": -142.598388671875,
|
| 168 |
+
"logps/rejected": -989.5442504882812,
|
| 169 |
+
"loss": 0.0079,
|
| 170 |
+
"rewards/accuracies": 1.0,
|
| 171 |
+
"rewards/chosen": -0.9251037836074829,
|
| 172 |
+
"rewards/margins": 7.951480865478516,
|
| 173 |
+
"rewards/rejected": -8.876585006713867,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.88,
|
| 178 |
+
"grad_norm": 0.010143597010992798,
|
| 179 |
+
"learning_rate": 9.828421365296023e-06,
|
| 180 |
+
"logits/chosen": -1.9893125295639038,
|
| 181 |
+
"logits/rejected": -2.527003526687622,
|
| 182 |
+
"logps/chosen": -127.56657409667969,
|
| 183 |
+
"logps/rejected": -1111.281005859375,
|
| 184 |
+
"loss": 0.0025,
|
| 185 |
+
"rewards/accuracies": 1.0,
|
| 186 |
+
"rewards/chosen": -0.7765494585037231,
|
| 187 |
+
"rewards/margins": 9.32172679901123,
|
| 188 |
+
"rewards/rejected": -10.098276138305664,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.96,
|
| 193 |
+
"grad_norm": 0.005924828079353313,
|
| 194 |
+
"learning_rate": 9.748325406443647e-06,
|
| 195 |
+
"logits/chosen": -1.330843448638916,
|
| 196 |
+
"logits/rejected": -2.0016732215881348,
|
| 197 |
+
"logps/chosen": -80.08639526367188,
|
| 198 |
+
"logps/rejected": -1047.385009765625,
|
| 199 |
+
"loss": 0.0004,
|
| 200 |
+
"rewards/accuracies": 1.0,
|
| 201 |
+
"rewards/chosen": -0.33333706855773926,
|
| 202 |
+
"rewards/margins": 9.099153518676758,
|
| 203 |
+
"rewards/rejected": -9.43249225616455,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.04,
|
| 208 |
+
"grad_norm": 0.006907045924266673,
|
| 209 |
+
"learning_rate": 9.653395576739504e-06,
|
| 210 |
+
"logits/chosen": -1.3165078163146973,
|
| 211 |
+
"logits/rejected": -1.999441385269165,
|
| 212 |
+
"logps/chosen": -142.83404541015625,
|
| 213 |
+
"logps/rejected": -1252.3265380859375,
|
| 214 |
+
"loss": 0.004,
|
| 215 |
+
"rewards/accuracies": 1.0,
|
| 216 |
+
"rewards/chosen": -0.9162995219230652,
|
| 217 |
+
"rewards/margins": 10.60672378540039,
|
| 218 |
+
"rewards/rejected": -11.52302360534668,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.12,
|
| 223 |
+
"grad_norm": 0.11065788109459221,
|
| 224 |
+
"learning_rate": 9.543928439016445e-06,
|
| 225 |
+
"logits/chosen": -2.337254047393799,
|
| 226 |
+
"logits/rejected": -2.8029887676239014,
|
| 227 |
+
"logps/chosen": -141.67652893066406,
|
| 228 |
+
"logps/rejected": -1141.8990478515625,
|
| 229 |
+
"loss": 0.0013,
|
| 230 |
+
"rewards/accuracies": 1.0,
|
| 231 |
+
"rewards/chosen": -0.9172345399856567,
|
| 232 |
+
"rewards/margins": 9.48987102508545,
|
| 233 |
+
"rewards/rejected": -10.407105445861816,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.2,
|
| 238 |
+
"grad_norm": 0.08367928566939553,
|
| 239 |
+
"learning_rate": 9.42026597097071e-06,
|
| 240 |
+
"logits/chosen": -2.5381298065185547,
|
| 241 |
+
"logits/rejected": -3.0234386920928955,
|
| 242 |
+
"logps/chosen": -162.30075073242188,
|
| 243 |
+
"logps/rejected": -1239.561767578125,
|
| 244 |
+
"loss": 0.0006,
|
| 245 |
+
"rewards/accuracies": 1.0,
|
| 246 |
+
"rewards/chosen": -1.1074072122573853,
|
| 247 |
+
"rewards/margins": 10.28199291229248,
|
| 248 |
+
"rewards/rejected": -11.389399528503418,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.28,
|
| 253 |
+
"grad_norm": 0.0036761092850006146,
|
| 254 |
+
"learning_rate": 9.282794496816244e-06,
|
| 255 |
+
"logits/chosen": -2.3002023696899414,
|
| 256 |
+
"logits/rejected": -2.8048605918884277,
|
| 257 |
+
"logps/chosen": -136.76681518554688,
|
| 258 |
+
"logps/rejected": -1181.3677978515625,
|
| 259 |
+
"loss": 0.0008,
|
| 260 |
+
"rewards/accuracies": 1.0,
|
| 261 |
+
"rewards/chosen": -0.8667460680007935,
|
| 262 |
+
"rewards/margins": 9.932957649230957,
|
| 263 |
+
"rewards/rejected": -10.799702644348145,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.3599999999999999,
|
| 268 |
+
"grad_norm": 0.0021196612287394884,
|
| 269 |
+
"learning_rate": 9.131943480399531e-06,
|
| 270 |
+
"logits/chosen": -2.4460995197296143,
|
| 271 |
+
"logits/rejected": -2.9539546966552734,
|
| 272 |
+
"logps/chosen": -154.1061553955078,
|
| 273 |
+
"logps/rejected": -1166.7919921875,
|
| 274 |
+
"loss": 0.0017,
|
| 275 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 276 |
+
"rewards/chosen": -1.0605894327163696,
|
| 277 |
+
"rewards/margins": 9.586730003356934,
|
| 278 |
+
"rewards/rejected": -10.647318840026855,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.44,
|
| 283 |
+
"grad_norm": 0.004351477647807626,
|
| 284 |
+
"learning_rate": 8.968184183545285e-06,
|
| 285 |
+
"logits/chosen": -1.3524295091629028,
|
| 286 |
+
"logits/rejected": -2.148332357406616,
|
| 287 |
+
"logps/chosen": -146.11656188964844,
|
| 288 |
+
"logps/rejected": -1293.1431884765625,
|
| 289 |
+
"loss": 0.0002,
|
| 290 |
+
"rewards/accuracies": 1.0,
|
| 291 |
+
"rewards/chosen": -0.959304928779602,
|
| 292 |
+
"rewards/margins": 10.965489387512207,
|
| 293 |
+
"rewards/rejected": -11.92479419708252,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.52,
|
| 298 |
+
"grad_norm": 0.004066031914054124,
|
| 299 |
+
"learning_rate": 8.792028193824364e-06,
|
| 300 |
+
"logits/chosen": -0.9521867036819458,
|
| 301 |
+
"logits/rejected": -1.7849302291870117,
|
| 302 |
+
"logps/chosen": -146.82675170898438,
|
| 303 |
+
"logps/rejected": -1224.1842041015625,
|
| 304 |
+
"loss": 0.0011,
|
| 305 |
+
"rewards/accuracies": 1.0,
|
| 306 |
+
"rewards/chosen": -0.9709192514419556,
|
| 307 |
+
"rewards/margins": 10.264577865600586,
|
| 308 |
+
"rewards/rejected": -11.235495567321777,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.6,
|
| 313 |
+
"grad_norm": 2.8254638303864446,
|
| 314 |
+
"learning_rate": 8.604025826343167e-06,
|
| 315 |
+
"logits/chosen": -1.4731212854385376,
|
| 316 |
+
"logits/rejected": -2.2158398628234863,
|
| 317 |
+
"logps/chosen": -166.7774200439453,
|
| 318 |
+
"logps/rejected": -1286.593505859375,
|
| 319 |
+
"loss": 0.0007,
|
| 320 |
+
"rewards/accuracies": 1.0,
|
| 321 |
+
"rewards/chosen": -1.1668568849563599,
|
| 322 |
+
"rewards/margins": 10.681219100952148,
|
| 323 |
+
"rewards/rejected": -11.848076820373535,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.6800000000000002,
|
| 328 |
+
"grad_norm": 0.027929253424062123,
|
| 329 |
+
"learning_rate": 8.404764404547404e-06,
|
| 330 |
+
"logits/chosen": -1.5285543203353882,
|
| 331 |
+
"logits/rejected": -2.207693099975586,
|
| 332 |
+
"logps/chosen": -145.4683837890625,
|
| 333 |
+
"logps/rejected": -1240.819091796875,
|
| 334 |
+
"loss": 0.0003,
|
| 335 |
+
"rewards/accuracies": 1.0,
|
| 336 |
+
"rewards/chosen": -0.9586501121520996,
|
| 337 |
+
"rewards/margins": 10.435168266296387,
|
| 338 |
+
"rewards/rejected": -11.393818855285645,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.76,
|
| 343 |
+
"grad_norm": 0.00891305422535769,
|
| 344 |
+
"learning_rate": 8.194866425410984e-06,
|
| 345 |
+
"logits/chosen": -1.2906019687652588,
|
| 346 |
+
"logits/rejected": -1.9918200969696045,
|
| 347 |
+
"logps/chosen": -109.07038879394531,
|
| 348 |
+
"logps/rejected": -1155.131591796875,
|
| 349 |
+
"loss": 0.0002,
|
| 350 |
+
"rewards/accuracies": 1.0,
|
| 351 |
+
"rewards/chosen": -0.6084994077682495,
|
| 352 |
+
"rewards/margins": 9.912607192993164,
|
| 353 |
+
"rewards/rejected": -10.521106719970703,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.8399999999999999,
|
| 358 |
+
"grad_norm": 0.0018316293380985671,
|
| 359 |
+
"learning_rate": 7.974987614742066e-06,
|
| 360 |
+
"logits/chosen": -0.9763646125793457,
|
| 361 |
+
"logits/rejected": -1.7541002035140991,
|
| 362 |
+
"logps/chosen": -96.57362365722656,
|
| 363 |
+
"logps/rejected": -1161.369384765625,
|
| 364 |
+
"loss": 0.0004,
|
| 365 |
+
"rewards/accuracies": 1.0,
|
| 366 |
+
"rewards/chosen": -0.4819282591342926,
|
| 367 |
+
"rewards/margins": 10.10207462310791,
|
| 368 |
+
"rewards/rejected": -10.584001541137695,
|
| 369 |
+
"step": 230
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.92,
|
| 373 |
+
"grad_norm": 0.003080239253269156,
|
| 374 |
+
"learning_rate": 7.745814878681516e-06,
|
| 375 |
+
"logits/chosen": -1.5342729091644287,
|
| 376 |
+
"logits/rejected": -2.221407175064087,
|
| 377 |
+
"logps/chosen": -136.1558074951172,
|
| 378 |
+
"logps/rejected": -1264.690673828125,
|
| 379 |
+
"loss": 0.0047,
|
| 380 |
+
"rewards/accuracies": 1.0,
|
| 381 |
+
"rewards/chosen": -0.8548258543014526,
|
| 382 |
+
"rewards/margins": 10.780150413513184,
|
| 383 |
+
"rewards/rejected": -11.634977340698242,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.0,
|
| 388 |
+
"grad_norm": 0.029482148279890704,
|
| 389 |
+
"learning_rate": 7.50806415779332e-06,
|
| 390 |
+
"logits/chosen": -2.258669853210449,
|
| 391 |
+
"logits/rejected": -2.8217015266418457,
|
| 392 |
+
"logps/chosen": -140.51853942871094,
|
| 393 |
+
"logps/rejected": -1197.581787109375,
|
| 394 |
+
"loss": 0.0003,
|
| 395 |
+
"rewards/accuracies": 1.0,
|
| 396 |
+
"rewards/chosen": -0.8967872858047485,
|
| 397 |
+
"rewards/margins": 10.075855255126953,
|
| 398 |
+
"rewards/rejected": -10.97264289855957,
|
| 399 |
+
"step": 250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.08,
|
| 403 |
+
"grad_norm": 0.020782306025551353,
|
| 404 |
+
"learning_rate": 7.262478190450834e-06,
|
| 405 |
+
"logits/chosen": -2.3182389736175537,
|
| 406 |
+
"logits/rejected": -2.92055082321167,
|
| 407 |
+
"logps/chosen": -120.69600677490234,
|
| 408 |
+
"logps/rejected": -1217.408447265625,
|
| 409 |
+
"loss": 0.0003,
|
| 410 |
+
"rewards/accuracies": 1.0,
|
| 411 |
+
"rewards/chosen": -0.7067109942436218,
|
| 412 |
+
"rewards/margins": 10.442468643188477,
|
| 413 |
+
"rewards/rejected": -11.14918041229248,
|
| 414 |
+
"step": 260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.16,
|
| 418 |
+
"grad_norm": 0.014378242296594932,
|
| 419 |
+
"learning_rate": 7.0098241925061215e-06,
|
| 420 |
+
"logits/chosen": -2.2274346351623535,
|
| 421 |
+
"logits/rejected": -2.8495638370513916,
|
| 422 |
+
"logps/chosen": -143.10069274902344,
|
| 423 |
+
"logps/rejected": -1256.431640625,
|
| 424 |
+
"loss": 0.0001,
|
| 425 |
+
"rewards/accuracies": 1.0,
|
| 426 |
+
"rewards/chosen": -0.919533371925354,
|
| 427 |
+
"rewards/margins": 10.646241188049316,
|
| 428 |
+
"rewards/rejected": -11.565774917602539,
|
| 429 |
+
"step": 270
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.24,
|
| 433 |
+
"grad_norm": 0.0011703826093326713,
|
| 434 |
+
"learning_rate": 6.750891460491093e-06,
|
| 435 |
+
"logits/chosen": -1.953466773033142,
|
| 436 |
+
"logits/rejected": -2.550438642501831,
|
| 437 |
+
"logps/chosen": -119.64122009277344,
|
| 438 |
+
"logps/rejected": -1253.8546142578125,
|
| 439 |
+
"loss": 0.0001,
|
| 440 |
+
"rewards/accuracies": 1.0,
|
| 441 |
+
"rewards/chosen": -0.6891164779663086,
|
| 442 |
+
"rewards/margins": 10.844141960144043,
|
| 443 |
+
"rewards/rejected": -11.533258438110352,
|
| 444 |
+
"step": 280
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.32,
|
| 448 |
+
"grad_norm": 0.016444776166175196,
|
| 449 |
+
"learning_rate": 6.486488905838143e-06,
|
| 450 |
+
"logits/chosen": -1.8186432123184204,
|
| 451 |
+
"logits/rejected": -2.431183338165283,
|
| 452 |
+
"logps/chosen": -114.60746002197266,
|
| 453 |
+
"logps/rejected": -1220.325927734375,
|
| 454 |
+
"loss": 0.0001,
|
| 455 |
+
"rewards/accuracies": 1.0,
|
| 456 |
+
"rewards/chosen": -0.6494568586349487,
|
| 457 |
+
"rewards/margins": 10.547185897827148,
|
| 458 |
+
"rewards/rejected": -11.196642875671387,
|
| 459 |
+
"step": 290
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.4,
|
| 463 |
+
"grad_norm": 0.0005496505009132721,
|
| 464 |
+
"learning_rate": 6.2174425278234115e-06,
|
| 465 |
+
"logits/chosen": -1.7246840000152588,
|
| 466 |
+
"logits/rejected": -2.2873055934906006,
|
| 467 |
+
"logps/chosen": -108.48016357421875,
|
| 468 |
+
"logps/rejected": -1233.37744140625,
|
| 469 |
+
"loss": 0.0001,
|
| 470 |
+
"rewards/accuracies": 1.0,
|
| 471 |
+
"rewards/chosen": -0.5782238245010376,
|
| 472 |
+
"rewards/margins": 10.742240905761719,
|
| 473 |
+
"rewards/rejected": -11.320464134216309,
|
| 474 |
+
"step": 300
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 2.48,
|
| 478 |
+
"grad_norm": 0.0006580056763243573,
|
| 479 |
+
"learning_rate": 5.944592833127253e-06,
|
| 480 |
+
"logits/chosen": -1.8141095638275146,
|
| 481 |
+
"logits/rejected": -2.429810047149658,
|
| 482 |
+
"logps/chosen": -119.1697006225586,
|
| 483 |
+
"logps/rejected": -1326.865966796875,
|
| 484 |
+
"loss": 0.0002,
|
| 485 |
+
"rewards/accuracies": 1.0,
|
| 486 |
+
"rewards/chosen": -0.6833556890487671,
|
| 487 |
+
"rewards/margins": 11.57683277130127,
|
| 488 |
+
"rewards/rejected": -12.260188102722168,
|
| 489 |
+
"step": 310
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.56,
|
| 493 |
+
"grad_norm": 0.0011989998803146951,
|
| 494 |
+
"learning_rate": 5.668792210073255e-06,
|
| 495 |
+
"logits/chosen": -1.7204673290252686,
|
| 496 |
+
"logits/rejected": -2.3327724933624268,
|
| 497 |
+
"logps/chosen": -103.1290512084961,
|
| 498 |
+
"logps/rejected": -1265.806640625,
|
| 499 |
+
"loss": 0.0001,
|
| 500 |
+
"rewards/accuracies": 1.0,
|
| 501 |
+
"rewards/chosen": -0.5410266518592834,
|
| 502 |
+
"rewards/margins": 11.101877212524414,
|
| 503 |
+
"rewards/rejected": -11.642904281616211,
|
| 504 |
+
"step": 320
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 2.64,
|
| 508 |
+
"grad_norm": 0.01089396369261857,
|
| 509 |
+
"learning_rate": 5.39090226574877e-06,
|
| 510 |
+
"logits/chosen": -1.52898371219635,
|
| 511 |
+
"logits/rejected": -2.140092134475708,
|
| 512 |
+
"logps/chosen": -91.94349670410156,
|
| 513 |
+
"logps/rejected": -1207.033447265625,
|
| 514 |
+
"loss": 0.0001,
|
| 515 |
+
"rewards/accuracies": 1.0,
|
| 516 |
+
"rewards/chosen": -0.43181926012039185,
|
| 517 |
+
"rewards/margins": 10.611051559448242,
|
| 518 |
+
"rewards/rejected": -11.042871475219727,
|
| 519 |
+
"step": 330
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.7199999999999998,
|
| 523 |
+
"grad_norm": 0.0015016734818494304,
|
| 524 |
+
"learning_rate": 5.111791134325793e-06,
|
| 525 |
+
"logits/chosen": -1.6063101291656494,
|
| 526 |
+
"logits/rejected": -2.3003313541412354,
|
| 527 |
+
"logps/chosen": -126.36214447021484,
|
| 528 |
+
"logps/rejected": -1331.2056884765625,
|
| 529 |
+
"loss": 0.0001,
|
| 530 |
+
"rewards/accuracies": 1.0,
|
| 531 |
+
"rewards/chosen": -0.7603046298027039,
|
| 532 |
+
"rewards/margins": 11.541264533996582,
|
| 533 |
+
"rewards/rejected": -12.301568984985352,
|
| 534 |
+
"step": 340
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.8,
|
| 538 |
+
"grad_norm": 0.044167012422386985,
|
| 539 |
+
"learning_rate": 4.832330764991131e-06,
|
| 540 |
+
"logits/chosen": -1.6678760051727295,
|
| 541 |
+
"logits/rejected": -2.332516670227051,
|
| 542 |
+
"logps/chosen": -119.7556381225586,
|
| 543 |
+
"logps/rejected": -1301.3248291015625,
|
| 544 |
+
"loss": 0.0004,
|
| 545 |
+
"rewards/accuracies": 1.0,
|
| 546 |
+
"rewards/chosen": -0.6866098642349243,
|
| 547 |
+
"rewards/margins": 11.313838958740234,
|
| 548 |
+
"rewards/rejected": -12.000448226928711,
|
| 549 |
+
"step": 350
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.88,
|
| 553 |
+
"grad_norm": 0.0005385533629396381,
|
| 554 |
+
"learning_rate": 4.553394197958339e-06,
|
| 555 |
+
"logits/chosen": -1.9416593313217163,
|
| 556 |
+
"logits/rejected": -2.6904163360595703,
|
| 557 |
+
"logps/chosen": -121.2122802734375,
|
| 558 |
+
"logps/rejected": -1295.510498046875,
|
| 559 |
+
"loss": 0.0001,
|
| 560 |
+
"rewards/accuracies": 1.0,
|
| 561 |
+
"rewards/chosen": -0.7095136642456055,
|
| 562 |
+
"rewards/margins": 11.224905014038086,
|
| 563 |
+
"rewards/rejected": -11.934419631958008,
|
| 564 |
+
"step": 360
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.96,
|
| 568 |
+
"grad_norm": 0.0006203949257092619,
|
| 569 |
+
"learning_rate": 4.275852837071309e-06,
|
| 570 |
+
"logits/chosen": -1.8644075393676758,
|
| 571 |
+
"logits/rejected": -2.7058677673339844,
|
| 572 |
+
"logps/chosen": -114.23481750488281,
|
| 573 |
+
"logps/rejected": -1273.7774658203125,
|
| 574 |
+
"loss": 0.0014,
|
| 575 |
+
"rewards/accuracies": 1.0,
|
| 576 |
+
"rewards/chosen": -0.6469079256057739,
|
| 577 |
+
"rewards/margins": 11.084717750549316,
|
| 578 |
+
"rewards/rejected": -11.731626510620117,
|
| 579 |
+
"step": 370
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.04,
|
| 583 |
+
"grad_norm": 0.0018701198703410247,
|
| 584 |
+
"learning_rate": 4.000573727519868e-06,
|
| 585 |
+
"logits/chosen": -1.6911453008651733,
|
| 586 |
+
"logits/rejected": -2.5057501792907715,
|
| 587 |
+
"logps/chosen": -115.5822982788086,
|
| 588 |
+
"logps/rejected": -1267.122314453125,
|
| 589 |
+
"loss": 0.0001,
|
| 590 |
+
"rewards/accuracies": 1.0,
|
| 591 |
+
"rewards/chosen": -0.6596534848213196,
|
| 592 |
+
"rewards/margins": 11.005937576293945,
|
| 593 |
+
"rewards/rejected": -11.665590286254883,
|
| 594 |
+
"step": 380
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.12,
|
| 598 |
+
"grad_norm": 0.004226077469779532,
|
| 599 |
+
"learning_rate": 3.7284168471719527e-06,
|
| 600 |
+
"logits/chosen": -1.646296501159668,
|
| 601 |
+
"logits/rejected": -2.4523534774780273,
|
| 602 |
+
"logps/chosen": -110.4627914428711,
|
| 603 |
+
"logps/rejected": -1280.544921875,
|
| 604 |
+
"loss": 0.0003,
|
| 605 |
+
"rewards/accuracies": 1.0,
|
| 606 |
+
"rewards/chosen": -0.6036595106124878,
|
| 607 |
+
"rewards/margins": 11.18139934539795,
|
| 608 |
+
"rewards/rejected": -11.785058975219727,
|
| 609 |
+
"step": 390
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.2,
|
| 613 |
+
"grad_norm": 0.0006904055330191614,
|
| 614 |
+
"learning_rate": 3.4602324199842026e-06,
|
| 615 |
+
"logits/chosen": -1.7251287698745728,
|
| 616 |
+
"logits/rejected": -2.5812900066375732,
|
| 617 |
+
"logps/chosen": -138.866943359375,
|
| 618 |
+
"logps/rejected": -1301.4326171875,
|
| 619 |
+
"loss": 0.0001,
|
| 620 |
+
"rewards/accuracies": 1.0,
|
| 621 |
+
"rewards/chosen": -0.8702784776687622,
|
| 622 |
+
"rewards/margins": 11.145977973937988,
|
| 623 |
+
"rewards/rejected": -12.016257286071777,
|
| 624 |
+
"step": 400
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.2800000000000002,
|
| 628 |
+
"grad_norm": 0.0026124260205581465,
|
| 629 |
+
"learning_rate": 3.1968582598840234e-06,
|
| 630 |
+
"logits/chosen": -1.5730335712432861,
|
| 631 |
+
"logits/rejected": -2.383654832839966,
|
| 632 |
+
"logps/chosen": -109.4728012084961,
|
| 633 |
+
"logps/rejected": -1262.404541015625,
|
| 634 |
+
"loss": 0.0,
|
| 635 |
+
"rewards/accuracies": 1.0,
|
| 636 |
+
"rewards/chosen": -0.6025822758674622,
|
| 637 |
+
"rewards/margins": 11.004800796508789,
|
| 638 |
+
"rewards/rejected": -11.607383728027344,
|
| 639 |
+
"step": 410
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.36,
|
| 643 |
+
"grad_norm": 0.016129813594014157,
|
| 644 |
+
"learning_rate": 2.9391171534208185e-06,
|
| 645 |
+
"logits/chosen": -1.7119560241699219,
|
| 646 |
+
"logits/rejected": -2.51015043258667,
|
| 647 |
+
"logps/chosen": -122.90666198730469,
|
| 648 |
+
"logps/rejected": -1313.09228515625,
|
| 649 |
+
"loss": 0.0001,
|
| 650 |
+
"rewards/accuracies": 1.0,
|
| 651 |
+
"rewards/chosen": -0.7348455190658569,
|
| 652 |
+
"rewards/margins": 11.394600868225098,
|
| 653 |
+
"rewards/rejected": -12.129446983337402,
|
| 654 |
+
"step": 420
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.44,
|
| 658 |
+
"grad_norm": 0.0024401290515470016,
|
| 659 |
+
"learning_rate": 2.6878142893630904e-06,
|
| 660 |
+
"logits/chosen": -1.6607134342193604,
|
| 661 |
+
"logits/rejected": -2.4547019004821777,
|
| 662 |
+
"logps/chosen": -137.77719116210938,
|
| 663 |
+
"logps/rejected": -1349.2073974609375,
|
| 664 |
+
"loss": 0.0001,
|
| 665 |
+
"rewards/accuracies": 1.0,
|
| 666 |
+
"rewards/chosen": -0.8648022413253784,
|
| 667 |
+
"rewards/margins": 11.622838020324707,
|
| 668 |
+
"rewards/rejected": -12.487640380859375,
|
| 669 |
+
"step": 430
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 3.52,
|
| 673 |
+
"grad_norm": 0.001375330125575793,
|
| 674 |
+
"learning_rate": 2.4437347432713838e-06,
|
| 675 |
+
"logits/chosen": -1.8045127391815186,
|
| 676 |
+
"logits/rejected": -2.586901903152466,
|
| 677 |
+
"logps/chosen": -156.62245178222656,
|
| 678 |
+
"logps/rejected": -1351.8687744140625,
|
| 679 |
+
"loss": 0.0001,
|
| 680 |
+
"rewards/accuracies": 1.0,
|
| 681 |
+
"rewards/chosen": -1.0343620777130127,
|
| 682 |
+
"rewards/margins": 11.499258995056152,
|
| 683 |
+
"rewards/rejected": -12.533620834350586,
|
| 684 |
+
"step": 440
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.6,
|
| 688 |
+
"grad_norm": 0.0016289824782741704,
|
| 689 |
+
"learning_rate": 2.207641024905322e-06,
|
| 690 |
+
"logits/chosen": -1.6454054117202759,
|
| 691 |
+
"logits/rejected": -2.4403579235076904,
|
| 692 |
+
"logps/chosen": -108.79969787597656,
|
| 693 |
+
"logps/rejected": -1255.9583740234375,
|
| 694 |
+
"loss": 0.0,
|
| 695 |
+
"rewards/accuracies": 1.0,
|
| 696 |
+
"rewards/chosen": -0.5948039293289185,
|
| 697 |
+
"rewards/margins": 10.957615852355957,
|
| 698 |
+
"rewards/rejected": -11.552419662475586,
|
| 699 |
+
"step": 450
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 3.68,
|
| 703 |
+
"grad_norm": 0.0003909667399708806,
|
| 704 |
+
"learning_rate": 1.9802706961266936e-06,
|
| 705 |
+
"logits/chosen": -1.5957154035568237,
|
| 706 |
+
"logits/rejected": -2.4980740547180176,
|
| 707 |
+
"logps/chosen": -124.23823547363281,
|
| 708 |
+
"logps/rejected": -1321.4207763671875,
|
| 709 |
+
"loss": 0.0002,
|
| 710 |
+
"rewards/accuracies": 1.0,
|
| 711 |
+
"rewards/chosen": -0.7479596138000488,
|
| 712 |
+
"rewards/margins": 11.454780578613281,
|
| 713 |
+
"rewards/rejected": -12.202739715576172,
|
| 714 |
+
"step": 460
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 3.76,
|
| 718 |
+
"grad_norm": 0.0006870240069107201,
|
| 719 |
+
"learning_rate": 1.7623340667403089e-06,
|
| 720 |
+
"logits/chosen": -1.5271575450897217,
|
| 721 |
+
"logits/rejected": -2.3313746452331543,
|
| 722 |
+
"logps/chosen": -115.49822998046875,
|
| 723 |
+
"logps/rejected": -1287.483642578125,
|
| 724 |
+
"loss": 0.0001,
|
| 725 |
+
"rewards/accuracies": 1.0,
|
| 726 |
+
"rewards/chosen": -0.6483306288719177,
|
| 727 |
+
"rewards/margins": 11.213019371032715,
|
| 728 |
+
"rewards/rejected": -11.861350059509277,
|
| 729 |
+
"step": 470
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.84,
|
| 733 |
+
"grad_norm": 0.03279067348481457,
|
| 734 |
+
"learning_rate": 1.5545119754708682e-06,
|
| 735 |
+
"logits/chosen": -1.719366431236267,
|
| 736 |
+
"logits/rejected": -2.5721096992492676,
|
| 737 |
+
"logps/chosen": -141.73440551757812,
|
| 738 |
+
"logps/rejected": -1358.951416015625,
|
| 739 |
+
"loss": 0.0001,
|
| 740 |
+
"rewards/accuracies": 1.0,
|
| 741 |
+
"rewards/chosen": -0.9100968241691589,
|
| 742 |
+
"rewards/margins": 11.68384838104248,
|
| 743 |
+
"rewards/rejected": -12.59394645690918,
|
| 744 |
+
"step": 480
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.92,
|
| 748 |
+
"grad_norm": 0.004304506433089471,
|
| 749 |
+
"learning_rate": 1.3574536630081208e-06,
|
| 750 |
+
"logits/chosen": -1.5175102949142456,
|
| 751 |
+
"logits/rejected": -2.292325496673584,
|
| 752 |
+
"logps/chosen": -100.6703872680664,
|
| 753 |
+
"logps/rejected": -1273.519775390625,
|
| 754 |
+
"loss": 0.0,
|
| 755 |
+
"rewards/accuracies": 1.0,
|
| 756 |
+
"rewards/chosen": -0.5087094902992249,
|
| 757 |
+
"rewards/margins": 11.203500747680664,
|
| 758 |
+
"rewards/rejected": -11.71220874786377,
|
| 759 |
+
"step": 490
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.0,
|
| 763 |
+
"grad_norm": 0.0004662700999661582,
|
| 764 |
+
"learning_rate": 1.1717747437649657e-06,
|
| 765 |
+
"logits/chosen": -1.5407252311706543,
|
| 766 |
+
"logits/rejected": -2.3491063117980957,
|
| 767 |
+
"logps/chosen": -122.54182434082031,
|
| 768 |
+
"logps/rejected": -1335.7579345703125,
|
| 769 |
+
"loss": 0.0,
|
| 770 |
+
"rewards/accuracies": 1.0,
|
| 771 |
+
"rewards/chosen": -0.7224875092506409,
|
| 772 |
+
"rewards/margins": 11.629507064819336,
|
| 773 |
+
"rewards/rejected": -12.351993560791016,
|
| 774 |
+
"step": 500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 4.08,
|
| 778 |
+
"grad_norm": 0.0017439323111767312,
|
| 779 |
+
"learning_rate": 9.980552826847635e-07,
|
| 780 |
+
"logits/chosen": -1.6905838251113892,
|
| 781 |
+
"logits/rejected": -2.487473487854004,
|
| 782 |
+
"logps/chosen": -166.93624877929688,
|
| 783 |
+
"logps/rejected": -1409.1981201171875,
|
| 784 |
+
"loss": 0.0001,
|
| 785 |
+
"rewards/accuracies": 1.0,
|
| 786 |
+
"rewards/chosen": -1.1514984369277954,
|
| 787 |
+
"rewards/margins": 11.959856033325195,
|
| 788 |
+
"rewards/rejected": -13.111355781555176,
|
| 789 |
+
"step": 510
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 4.16,
|
| 793 |
+
"grad_norm": 0.0024902235420863005,
|
| 794 |
+
"learning_rate": 8.368379831059592e-07,
|
| 795 |
+
"logits/chosen": -1.5832383632659912,
|
| 796 |
+
"logits/rejected": -2.3622446060180664,
|
| 797 |
+
"logps/chosen": -130.75253295898438,
|
| 798 |
+
"logps/rejected": -1372.863037109375,
|
| 799 |
+
"loss": 0.0,
|
| 800 |
+
"rewards/accuracies": 1.0,
|
| 801 |
+
"rewards/chosen": -0.806990921497345,
|
| 802 |
+
"rewards/margins": 11.918740272521973,
|
| 803 |
+
"rewards/rejected": -12.725730895996094,
|
| 804 |
+
"step": 520
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 4.24,
|
| 808 |
+
"grad_norm": 0.0004030809278175782,
|
| 809 |
+
"learning_rate": 6.886264913451635e-07,
|
| 810 |
+
"logits/chosen": -1.5754905939102173,
|
| 811 |
+
"logits/rejected": -2.314927101135254,
|
| 812 |
+
"logps/chosen": -109.8852310180664,
|
| 813 |
+
"logps/rejected": -1284.387451171875,
|
| 814 |
+
"loss": 0.0,
|
| 815 |
+
"rewards/accuracies": 1.0,
|
| 816 |
+
"rewards/chosen": -0.6023492813110352,
|
| 817 |
+
"rewards/margins": 11.232891082763672,
|
| 818 |
+
"rewards/rejected": -11.835241317749023,
|
| 819 |
+
"step": 530
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.32,
|
| 823 |
+
"grad_norm": 0.014990693875413815,
|
| 824 |
+
"learning_rate": 5.538838232952104e-07,
|
| 825 |
+
"logits/chosen": -1.6247650384902954,
|
| 826 |
+
"logits/rejected": -2.395631790161133,
|
| 827 |
+
"logps/chosen": -112.5528335571289,
|
| 828 |
+
"logps/rejected": -1308.465087890625,
|
| 829 |
+
"loss": 0.0001,
|
| 830 |
+
"rewards/accuracies": 1.0,
|
| 831 |
+
"rewards/chosen": -0.6267929673194885,
|
| 832 |
+
"rewards/margins": 11.434622764587402,
|
| 833 |
+
"rewards/rejected": -12.06141471862793,
|
| 834 |
+
"step": 540
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 4.4,
|
| 838 |
+
"grad_norm": 0.0008931683311901964,
|
| 839 |
+
"learning_rate": 4.3303091795353024e-07,
|
| 840 |
+
"logits/chosen": -1.7327635288238525,
|
| 841 |
+
"logits/rejected": -2.521054744720459,
|
| 842 |
+
"logps/chosen": -193.12649536132812,
|
| 843 |
+
"logps/rejected": -1453.8939208984375,
|
| 844 |
+
"loss": 0.0,
|
| 845 |
+
"rewards/accuracies": 1.0,
|
| 846 |
+
"rewards/chosen": -1.3928953409194946,
|
| 847 |
+
"rewards/margins": 12.179082870483398,
|
| 848 |
+
"rewards/rejected": -13.571980476379395,
|
| 849 |
+
"step": 550
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.48,
|
| 853 |
+
"grad_norm": 0.0006090223103858115,
|
| 854 |
+
"learning_rate": 3.2644532239966444e-07,
|
| 855 |
+
"logits/chosen": -1.540307879447937,
|
| 856 |
+
"logits/rejected": -2.274130344390869,
|
| 857 |
+
"logps/chosen": -116.36529541015625,
|
| 858 |
+
"logps/rejected": -1331.5732421875,
|
| 859 |
+
"loss": 0.0001,
|
| 860 |
+
"rewards/accuracies": 1.0,
|
| 861 |
+
"rewards/chosen": -0.6516658663749695,
|
| 862 |
+
"rewards/margins": 11.651314735412598,
|
| 863 |
+
"rewards/rejected": -12.302980422973633,
|
| 864 |
+
"step": 560
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.5600000000000005,
|
| 868 |
+
"grad_norm": 0.0012600919102292293,
|
| 869 |
+
"learning_rate": 2.3446001233004333e-07,
|
| 870 |
+
"logits/chosen": -1.545082926750183,
|
| 871 |
+
"logits/rejected": -2.3070833683013916,
|
| 872 |
+
"logps/chosen": -118.43846130371094,
|
| 873 |
+
"logps/rejected": -1310.844482421875,
|
| 874 |
+
"loss": 0.0001,
|
| 875 |
+
"rewards/accuracies": 1.0,
|
| 876 |
+
"rewards/chosen": -0.6805299520492554,
|
| 877 |
+
"rewards/margins": 11.414255142211914,
|
| 878 |
+
"rewards/rejected": -12.094785690307617,
|
| 879 |
+
"step": 570
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 4.64,
|
| 883 |
+
"grad_norm": 0.0033895825817679913,
|
| 884 |
+
"learning_rate": 1.573623518347517e-07,
|
| 885 |
+
"logits/chosen": -1.3955062627792358,
|
| 886 |
+
"logits/rejected": -2.1139488220214844,
|
| 887 |
+
"logps/chosen": -95.18065643310547,
|
| 888 |
+
"logps/rejected": -1255.7984619140625,
|
| 889 |
+
"loss": 0.0,
|
| 890 |
+
"rewards/accuracies": 1.0,
|
| 891 |
+
"rewards/chosen": -0.46435871720314026,
|
| 892 |
+
"rewards/margins": 11.06248664855957,
|
| 893 |
+
"rewards/rejected": -11.52684497833252,
|
| 894 |
+
"step": 580
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.72,
|
| 898 |
+
"grad_norm": 0.0005277953893460651,
|
| 899 |
+
"learning_rate": 9.539319566590766e-08,
|
| 900 |
+
"logits/chosen": -1.4713037014007568,
|
| 901 |
+
"logits/rejected": -2.1673684120178223,
|
| 902 |
+
"logps/chosen": -103.91761779785156,
|
| 903 |
+
"logps/rejected": -1269.9681396484375,
|
| 904 |
+
"loss": 0.0,
|
| 905 |
+
"rewards/accuracies": 1.0,
|
| 906 |
+
"rewards/chosen": -0.5508452653884888,
|
| 907 |
+
"rewards/margins": 11.134601593017578,
|
| 908 |
+
"rewards/rejected": -11.685447692871094,
|
| 909 |
+
"step": 590
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 4.8,
|
| 913 |
+
"grad_norm": 0.0038828625096458437,
|
| 914 |
+
"learning_rate": 4.8746136802240716e-08,
|
| 915 |
+
"logits/chosen": -1.60639226436615,
|
| 916 |
+
"logits/rejected": -2.2908573150634766,
|
| 917 |
+
"logps/chosen": -111.3029556274414,
|
| 918 |
+
"logps/rejected": -1284.128173828125,
|
| 919 |
+
"loss": 0.0001,
|
| 920 |
+
"rewards/accuracies": 1.0,
|
| 921 |
+
"rewards/chosen": -0.6181883811950684,
|
| 922 |
+
"rewards/margins": 11.210031509399414,
|
| 923 |
+
"rewards/rejected": -11.82822036743164,
|
| 924 |
+
"step": 600
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 4.88,
|
| 928 |
+
"grad_norm": 0.0016152355239910333,
|
| 929 |
+
"learning_rate": 1.75669016604485e-08,
|
| 930 |
+
"logits/chosen": -1.5895562171936035,
|
| 931 |
+
"logits/rejected": -2.336843490600586,
|
| 932 |
+
"logps/chosen": -118.6875991821289,
|
| 933 |
+
"logps/rejected": -1293.158935546875,
|
| 934 |
+
"loss": 0.0,
|
| 935 |
+
"rewards/accuracies": 1.0,
|
| 936 |
+
"rewards/chosen": -0.6965440511703491,
|
| 937 |
+
"rewards/margins": 11.21881103515625,
|
| 938 |
+
"rewards/rejected": -11.915353775024414,
|
| 939 |
+
"step": 610
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 4.96,
|
| 943 |
+
"grad_norm": 0.0004233911082118195,
|
| 944 |
+
"learning_rate": 1.952894842735531e-09,
|
| 945 |
+
"logits/chosen": -1.5527576208114624,
|
| 946 |
+
"logits/rejected": -2.2691588401794434,
|
| 947 |
+
"logps/chosen": -98.73721313476562,
|
| 948 |
+
"logps/rejected": -1254.2822265625,
|
| 949 |
+
"loss": 0.0,
|
| 950 |
+
"rewards/accuracies": 1.0,
|
| 951 |
+
"rewards/chosen": -0.4978357255458832,
|
| 952 |
+
"rewards/margins": 11.023382186889648,
|
| 953 |
+
"rewards/rejected": -11.52121639251709,
|
| 954 |
+
"step": 620
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 5.0,
|
| 958 |
+
"step": 625,
|
| 959 |
+
"total_flos": 0.0,
|
| 960 |
+
"train_loss": 0.022511796173290348,
|
| 961 |
+
"train_runtime": 65461.2558,
|
| 962 |
+
"train_samples_per_second": 0.611,
|
| 963 |
+
"train_steps_per_second": 0.01
|
| 964 |
+
}
|
| 965 |
+
],
|
| 966 |
+
"logging_steps": 10,
|
| 967 |
+
"max_steps": 625,
|
| 968 |
+
"num_input_tokens_seen": 0,
|
| 969 |
+
"num_train_epochs": 5,
|
| 970 |
+
"save_steps": 1000,
|
| 971 |
+
"stateful_callbacks": {
|
| 972 |
+
"TrainerControl": {
|
| 973 |
+
"args": {
|
| 974 |
+
"should_epoch_stop": false,
|
| 975 |
+
"should_evaluate": false,
|
| 976 |
+
"should_log": false,
|
| 977 |
+
"should_save": true,
|
| 978 |
+
"should_training_stop": true
|
| 979 |
+
},
|
| 980 |
+
"attributes": {}
|
| 981 |
+
}
|
| 982 |
+
},
|
| 983 |
+
"total_flos": 0.0,
|
| 984 |
+
"train_batch_size": 4,
|
| 985 |
+
"trial_name": null,
|
| 986 |
+
"trial_params": null
|
| 987 |
+
}
|