Model save
Browse files- README.md +77 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +571 -0
README.md
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
tags:
|
| 4 |
+
- trl
|
| 5 |
+
- dpo
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
base_model: allenai/tulu-2-7b
|
| 8 |
+
model-index:
|
| 9 |
+
- name: tulu2-7b-cost-UI-both-5e-7
|
| 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 |
+
# tulu2-7b-cost-UI-both-5e-7
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [allenai/tulu-2-7b](https://huggingface.co/allenai/tulu-2-7b) on the None dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.6928
|
| 21 |
+
- Rewards/chosen: -0.0002
|
| 22 |
+
- Rewards/rejected: -0.0001
|
| 23 |
+
- Rewards/accuracies: 0.5159
|
| 24 |
+
- Rewards/margins: -0.0001
|
| 25 |
+
- Rewards/margins Max: 0.0190
|
| 26 |
+
- Rewards/margins Min: -0.0190
|
| 27 |
+
- Rewards/margins Std: 0.0125
|
| 28 |
+
- Logps/rejected: -338.8891
|
| 29 |
+
- Logps/chosen: -345.6230
|
| 30 |
+
- Logits/rejected: 0.8420
|
| 31 |
+
- Logits/chosen: 0.7124
|
| 32 |
+
|
| 33 |
+
## Model description
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Intended uses & limitations
|
| 38 |
+
|
| 39 |
+
More information needed
|
| 40 |
+
|
| 41 |
+
## Training and evaluation data
|
| 42 |
+
|
| 43 |
+
More information needed
|
| 44 |
+
|
| 45 |
+
## Training procedure
|
| 46 |
+
|
| 47 |
+
### Training hyperparameters
|
| 48 |
+
|
| 49 |
+
The following hyperparameters were used during training:
|
| 50 |
+
- learning_rate: 5e-07
|
| 51 |
+
- train_batch_size: 2
|
| 52 |
+
- eval_batch_size: 8
|
| 53 |
+
- seed: 42
|
| 54 |
+
- distributed_type: multi-GPU
|
| 55 |
+
- num_devices: 8
|
| 56 |
+
- gradient_accumulation_steps: 4
|
| 57 |
+
- total_train_batch_size: 64
|
| 58 |
+
- total_eval_batch_size: 64
|
| 59 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 60 |
+
- lr_scheduler_type: cosine
|
| 61 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 62 |
+
- num_epochs: 1
|
| 63 |
+
|
| 64 |
+
### Training results
|
| 65 |
+
|
| 66 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Rewards/margins Max | Rewards/margins Min | Rewards/margins Std | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 67 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:-------------------:|:-------------------:|:-------------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 68 |
+
| 0.6764 | 1.0 | 289 | 0.6928 | -0.0002 | -0.0001 | 0.5159 | -0.0001 | 0.0190 | -0.0190 | 0.0125 | -338.8891 | -345.6230 | 0.8420 | 0.7124 |
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
### Framework versions
|
| 72 |
+
|
| 73 |
+
- PEFT 0.7.1
|
| 74 |
+
- Transformers 4.39.0.dev0
|
| 75 |
+
- Pytorch 2.1.2+cu121
|
| 76 |
+
- Datasets 2.14.6
|
| 77 |
+
- Tokenizers 0.15.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 639692768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86739c0322f6b1ea5b274578df67afec4d4aa82c2927f6a3add22ee23935c015
|
| 3 |
size 639692768
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.6824996727148142,
|
| 4 |
+
"train_runtime": 3006.8067,
|
| 5 |
+
"train_samples": 18509,
|
| 6 |
+
"train_samples_per_second": 6.156,
|
| 7 |
+
"train_steps_per_second": 0.096
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.6824996727148142,
|
| 4 |
+
"train_runtime": 3006.8067,
|
| 5 |
+
"train_samples": 18509,
|
| 6 |
+
"train_samples_per_second": 6.156,
|
| 7 |
+
"train_steps_per_second": 0.096
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,571 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9991356957649092,
|
| 5 |
+
"eval_steps": 100,
|
| 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": 0.0,
|
| 13 |
+
"grad_norm": 0.2373046875,
|
| 14 |
+
"learning_rate": 1.7241379310344825e-08,
|
| 15 |
+
"logits/chosen": -0.1198696494102478,
|
| 16 |
+
"logits/rejected": 0.2811677157878876,
|
| 17 |
+
"logps/chosen": -217.16847229003906,
|
| 18 |
+
"logps/rejected": -155.17271423339844,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/margins_max": 0.0,
|
| 24 |
+
"rewards/margins_min": 0.0,
|
| 25 |
+
"rewards/margins_std": 0.0,
|
| 26 |
+
"rewards/rejected": 0.0,
|
| 27 |
+
"step": 1
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.03,
|
| 31 |
+
"grad_norm": 0.248046875,
|
| 32 |
+
"learning_rate": 1.7241379310344828e-07,
|
| 33 |
+
"logits/chosen": -0.13907119631767273,
|
| 34 |
+
"logits/rejected": 0.10448222607374191,
|
| 35 |
+
"logps/chosen": -224.61839294433594,
|
| 36 |
+
"logps/rejected": -217.18283081054688,
|
| 37 |
+
"loss": 0.6932,
|
| 38 |
+
"rewards/accuracies": 0.4027777910232544,
|
| 39 |
+
"rewards/chosen": -0.00020672479877248406,
|
| 40 |
+
"rewards/margins": -0.0004321257583796978,
|
| 41 |
+
"rewards/margins_max": 0.0018353578634560108,
|
| 42 |
+
"rewards/margins_min": -0.0026996093802154064,
|
| 43 |
+
"rewards/margins_std": 0.0032067059073597193,
|
| 44 |
+
"rewards/rejected": 0.00022540091595146805,
|
| 45 |
+
"step": 10
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07,
|
| 49 |
+
"grad_norm": 0.2353515625,
|
| 50 |
+
"learning_rate": 3.4482758620689656e-07,
|
| 51 |
+
"logits/chosen": -0.19345124065876007,
|
| 52 |
+
"logits/rejected": 0.05696944519877434,
|
| 53 |
+
"logps/chosen": -191.2423553466797,
|
| 54 |
+
"logps/rejected": -176.5889434814453,
|
| 55 |
+
"loss": 0.6929,
|
| 56 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 57 |
+
"rewards/chosen": -2.7165409846929833e-05,
|
| 58 |
+
"rewards/margins": 0.0008886352297849953,
|
| 59 |
+
"rewards/margins_max": 0.002847478026524186,
|
| 60 |
+
"rewards/margins_min": -0.0010702075669541955,
|
| 61 |
+
"rewards/margins_std": 0.0027702220249921083,
|
| 62 |
+
"rewards/rejected": -0.0009158005705103278,
|
| 63 |
+
"step": 20
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.1,
|
| 67 |
+
"grad_norm": 0.265625,
|
| 68 |
+
"learning_rate": 4.999817502139027e-07,
|
| 69 |
+
"logits/chosen": -0.19917765259742737,
|
| 70 |
+
"logits/rejected": 0.08285371214151382,
|
| 71 |
+
"logps/chosen": -247.91159057617188,
|
| 72 |
+
"logps/rejected": -218.81246948242188,
|
| 73 |
+
"loss": 0.6922,
|
| 74 |
+
"rewards/accuracies": 0.75,
|
| 75 |
+
"rewards/chosen": 0.0009048490901477635,
|
| 76 |
+
"rewards/margins": 0.0024331805761903524,
|
| 77 |
+
"rewards/margins_max": 0.005544544197618961,
|
| 78 |
+
"rewards/margins_min": -0.0006781836273148656,
|
| 79 |
+
"rewards/margins_std": 0.004400133155286312,
|
| 80 |
+
"rewards/rejected": -0.0015283313114196062,
|
| 81 |
+
"step": 30
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.14,
|
| 85 |
+
"grad_norm": 0.267578125,
|
| 86 |
+
"learning_rate": 4.977949980164773e-07,
|
| 87 |
+
"logits/chosen": -0.05700277164578438,
|
| 88 |
+
"logits/rejected": 0.11779887974262238,
|
| 89 |
+
"logps/chosen": -193.61355590820312,
|
| 90 |
+
"logps/rejected": -203.84695434570312,
|
| 91 |
+
"loss": 0.6914,
|
| 92 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 93 |
+
"rewards/chosen": 0.0017391880974173546,
|
| 94 |
+
"rewards/margins": 0.004240007139742374,
|
| 95 |
+
"rewards/margins_max": 0.008039236068725586,
|
| 96 |
+
"rewards/margins_min": 0.0004407777450978756,
|
| 97 |
+
"rewards/margins_std": 0.005372921004891396,
|
| 98 |
+
"rewards/rejected": -0.0025008185766637325,
|
| 99 |
+
"step": 40
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.17,
|
| 103 |
+
"grad_norm": 0.2314453125,
|
| 104 |
+
"learning_rate": 4.919948367622307e-07,
|
| 105 |
+
"logits/chosen": -0.1814556121826172,
|
| 106 |
+
"logits/rejected": 0.02914687618613243,
|
| 107 |
+
"logps/chosen": -220.93115234375,
|
| 108 |
+
"logps/rejected": -223.7554168701172,
|
| 109 |
+
"loss": 0.6902,
|
| 110 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 111 |
+
"rewards/chosen": 0.0025483686476945877,
|
| 112 |
+
"rewards/margins": 0.0058803861029446125,
|
| 113 |
+
"rewards/margins_max": 0.009798353537917137,
|
| 114 |
+
"rewards/margins_min": 0.0019624175038188696,
|
| 115 |
+
"rewards/margins_std": 0.005540843587368727,
|
| 116 |
+
"rewards/rejected": -0.003332017455250025,
|
| 117 |
+
"step": 50
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.21,
|
| 121 |
+
"grad_norm": 0.2275390625,
|
| 122 |
+
"learning_rate": 4.826658458630755e-07,
|
| 123 |
+
"logits/chosen": -0.17724668979644775,
|
| 124 |
+
"logits/rejected": 0.1624217927455902,
|
| 125 |
+
"logps/chosen": -226.7654571533203,
|
| 126 |
+
"logps/rejected": -178.1088104248047,
|
| 127 |
+
"loss": 0.6892,
|
| 128 |
+
"rewards/accuracies": 0.925000011920929,
|
| 129 |
+
"rewards/chosen": 0.004519807640463114,
|
| 130 |
+
"rewards/margins": 0.008648158982396126,
|
| 131 |
+
"rewards/margins_max": 0.013417336158454418,
|
| 132 |
+
"rewards/margins_min": 0.0038789804093539715,
|
| 133 |
+
"rewards/margins_std": 0.006744635757058859,
|
| 134 |
+
"rewards/rejected": -0.004128350876271725,
|
| 135 |
+
"step": 60
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.24,
|
| 139 |
+
"grad_norm": 0.248046875,
|
| 140 |
+
"learning_rate": 4.699440630133794e-07,
|
| 141 |
+
"logits/chosen": -0.15689274668693542,
|
| 142 |
+
"logits/rejected": 0.0655246376991272,
|
| 143 |
+
"logps/chosen": -205.931884765625,
|
| 144 |
+
"logps/rejected": -213.77328491210938,
|
| 145 |
+
"loss": 0.6879,
|
| 146 |
+
"rewards/accuracies": 0.862500011920929,
|
| 147 |
+
"rewards/chosen": 0.004349945578724146,
|
| 148 |
+
"rewards/margins": 0.010797133669257164,
|
| 149 |
+
"rewards/margins_max": 0.016608919948339462,
|
| 150 |
+
"rewards/margins_min": 0.004985347390174866,
|
| 151 |
+
"rewards/margins_std": 0.008219106122851372,
|
| 152 |
+
"rewards/rejected": -0.006447188556194305,
|
| 153 |
+
"step": 70
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.28,
|
| 157 |
+
"grad_norm": 0.267578125,
|
| 158 |
+
"learning_rate": 4.5401500045405117e-07,
|
| 159 |
+
"logits/chosen": -0.13238832354545593,
|
| 160 |
+
"logits/rejected": 0.1644008904695511,
|
| 161 |
+
"logps/chosen": -232.31631469726562,
|
| 162 |
+
"logps/rejected": -199.2275390625,
|
| 163 |
+
"loss": 0.6865,
|
| 164 |
+
"rewards/accuracies": 0.887499988079071,
|
| 165 |
+
"rewards/chosen": 0.004281439818441868,
|
| 166 |
+
"rewards/margins": 0.012644032016396523,
|
| 167 |
+
"rewards/margins_max": 0.01821981742978096,
|
| 168 |
+
"rewards/margins_min": 0.007068246603012085,
|
| 169 |
+
"rewards/margins_std": 0.007885349914431572,
|
| 170 |
+
"rewards/rejected": -0.00836259126663208,
|
| 171 |
+
"step": 80
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.31,
|
| 175 |
+
"grad_norm": 0.244140625,
|
| 176 |
+
"learning_rate": 4.3511093978633086e-07,
|
| 177 |
+
"logits/chosen": -0.2108915150165558,
|
| 178 |
+
"logits/rejected": -0.023733098059892654,
|
| 179 |
+
"logps/chosen": -223.54336547851562,
|
| 180 |
+
"logps/rejected": -231.7197265625,
|
| 181 |
+
"loss": 0.6859,
|
| 182 |
+
"rewards/accuracies": 0.887499988079071,
|
| 183 |
+
"rewards/chosen": 0.006567983888089657,
|
| 184 |
+
"rewards/margins": 0.016123134642839432,
|
| 185 |
+
"rewards/margins_max": 0.02399933896958828,
|
| 186 |
+
"rewards/margins_min": 0.008246931247413158,
|
| 187 |
+
"rewards/margins_std": 0.011138634756207466,
|
| 188 |
+
"rewards/rejected": -0.009555150754749775,
|
| 189 |
+
"step": 90
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.35,
|
| 193 |
+
"grad_norm": 0.2451171875,
|
| 194 |
+
"learning_rate": 4.135075447829911e-07,
|
| 195 |
+
"logits/chosen": -0.13402007520198822,
|
| 196 |
+
"logits/rejected": 0.23311862349510193,
|
| 197 |
+
"logps/chosen": -242.7605438232422,
|
| 198 |
+
"logps/rejected": -215.2969970703125,
|
| 199 |
+
"loss": 0.6846,
|
| 200 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 201 |
+
"rewards/chosen": 0.006707096938043833,
|
| 202 |
+
"rewards/margins": 0.020004788413643837,
|
| 203 |
+
"rewards/margins_max": 0.03050144948065281,
|
| 204 |
+
"rewards/margins_min": 0.009508123621344566,
|
| 205 |
+
"rewards/margins_std": 0.01484452374279499,
|
| 206 |
+
"rewards/rejected": -0.013297691941261292,
|
| 207 |
+
"step": 100
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.38,
|
| 211 |
+
"grad_norm": 0.240234375,
|
| 212 |
+
"learning_rate": 3.895198415897896e-07,
|
| 213 |
+
"logits/chosen": -0.10960109531879425,
|
| 214 |
+
"logits/rejected": 0.09310563653707504,
|
| 215 |
+
"logps/chosen": -215.31460571289062,
|
| 216 |
+
"logps/rejected": -208.97097778320312,
|
| 217 |
+
"loss": 0.684,
|
| 218 |
+
"rewards/accuracies": 0.887499988079071,
|
| 219 |
+
"rewards/chosen": 0.007017076015472412,
|
| 220 |
+
"rewards/margins": 0.01683773659169674,
|
| 221 |
+
"rewards/margins_max": 0.026084523648023605,
|
| 222 |
+
"rewards/margins_min": 0.007590950932353735,
|
| 223 |
+
"rewards/margins_std": 0.01307692937552929,
|
| 224 |
+
"rewards/rejected": -0.009820659644901752,
|
| 225 |
+
"step": 110
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.41,
|
| 229 |
+
"grad_norm": 0.2451171875,
|
| 230 |
+
"learning_rate": 3.634976249348867e-07,
|
| 231 |
+
"logits/chosen": -0.13226808607578278,
|
| 232 |
+
"logits/rejected": 0.15200194716453552,
|
| 233 |
+
"logps/chosen": -242.24966430664062,
|
| 234 |
+
"logps/rejected": -212.1212921142578,
|
| 235 |
+
"loss": 0.6826,
|
| 236 |
+
"rewards/accuracies": 0.824999988079071,
|
| 237 |
+
"rewards/chosen": 0.006220725364983082,
|
| 238 |
+
"rewards/margins": 0.01998782530426979,
|
| 239 |
+
"rewards/margins_max": 0.03047138825058937,
|
| 240 |
+
"rewards/margins_min": 0.009504261426627636,
|
| 241 |
+
"rewards/margins_std": 0.014825996942818165,
|
| 242 |
+
"rewards/rejected": -0.013767099007964134,
|
| 243 |
+
"step": 120
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.45,
|
| 247 |
+
"grad_norm": 0.2353515625,
|
| 248 |
+
"learning_rate": 3.358203573340396e-07,
|
| 249 |
+
"logits/chosen": -0.15161535143852234,
|
| 250 |
+
"logits/rejected": 0.06725052744150162,
|
| 251 |
+
"logps/chosen": -220.3600311279297,
|
| 252 |
+
"logps/rejected": -238.15267944335938,
|
| 253 |
+
"loss": 0.6817,
|
| 254 |
+
"rewards/accuracies": 0.875,
|
| 255 |
+
"rewards/chosen": 0.00594148738309741,
|
| 256 |
+
"rewards/margins": 0.02438289485871792,
|
| 257 |
+
"rewards/margins_max": 0.03674236685037613,
|
| 258 |
+
"rewards/margins_min": 0.012023425661027431,
|
| 259 |
+
"rewards/margins_std": 0.017478929832577705,
|
| 260 |
+
"rewards/rejected": -0.01844140887260437,
|
| 261 |
+
"step": 130
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.48,
|
| 265 |
+
"grad_norm": 0.2373046875,
|
| 266 |
+
"learning_rate": 3.0689163567264746e-07,
|
| 267 |
+
"logits/chosen": -0.1485133320093155,
|
| 268 |
+
"logits/rejected": 0.12535016238689423,
|
| 269 |
+
"logps/chosen": -246.51065063476562,
|
| 270 |
+
"logps/rejected": -233.1022186279297,
|
| 271 |
+
"loss": 0.6808,
|
| 272 |
+
"rewards/accuracies": 0.875,
|
| 273 |
+
"rewards/chosen": 0.009529241360723972,
|
| 274 |
+
"rewards/margins": 0.021075651049613953,
|
| 275 |
+
"rewards/margins_max": 0.03406635671854019,
|
| 276 |
+
"rewards/margins_min": 0.008084945380687714,
|
| 277 |
+
"rewards/margins_std": 0.01837163418531418,
|
| 278 |
+
"rewards/rejected": -0.011546410620212555,
|
| 279 |
+
"step": 140
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.52,
|
| 283 |
+
"grad_norm": 0.2470703125,
|
| 284 |
+
"learning_rate": 2.771333058543416e-07,
|
| 285 |
+
"logits/chosen": -0.0800982341170311,
|
| 286 |
+
"logits/rejected": 0.1670171320438385,
|
| 287 |
+
"logps/chosen": -229.39120483398438,
|
| 288 |
+
"logps/rejected": -211.8321990966797,
|
| 289 |
+
"loss": 0.6804,
|
| 290 |
+
"rewards/accuracies": 0.887499988079071,
|
| 291 |
+
"rewards/chosen": 0.010588416829705238,
|
| 292 |
+
"rewards/margins": 0.026508808135986328,
|
| 293 |
+
"rewards/margins_max": 0.043856311589479446,
|
| 294 |
+
"rewards/margins_min": 0.009161303751170635,
|
| 295 |
+
"rewards/margins_std": 0.024533074349164963,
|
| 296 |
+
"rewards/rejected": -0.01592039130628109,
|
| 297 |
+
"step": 150
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.55,
|
| 301 |
+
"grad_norm": 0.482421875,
|
| 302 |
+
"learning_rate": 2.469793113377957e-07,
|
| 303 |
+
"logits/chosen": -0.12969347834587097,
|
| 304 |
+
"logits/rejected": 0.2130148708820343,
|
| 305 |
+
"logps/chosen": -207.0413818359375,
|
| 306 |
+
"logps/rejected": -174.50192260742188,
|
| 307 |
+
"loss": 0.6789,
|
| 308 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 309 |
+
"rewards/chosen": 0.009559371508657932,
|
| 310 |
+
"rewards/margins": 0.02586180530488491,
|
| 311 |
+
"rewards/margins_max": 0.04229948669672012,
|
| 312 |
+
"rewards/margins_min": 0.009424128569662571,
|
| 313 |
+
"rewards/margins_std": 0.023246387019753456,
|
| 314 |
+
"rewards/rejected": -0.016302434727549553,
|
| 315 |
+
"step": 160
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.59,
|
| 319 |
+
"grad_norm": 0.2265625,
|
| 320 |
+
"learning_rate": 2.1686936526394318e-07,
|
| 321 |
+
"logits/chosen": -0.1643756926059723,
|
| 322 |
+
"logits/rejected": 0.09980125725269318,
|
| 323 |
+
"logps/chosen": -232.63479614257812,
|
| 324 |
+
"logps/rejected": -222.15707397460938,
|
| 325 |
+
"loss": 0.6785,
|
| 326 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 327 |
+
"rewards/chosen": 0.009239903651177883,
|
| 328 |
+
"rewards/margins": 0.031400762498378754,
|
| 329 |
+
"rewards/margins_max": 0.04692839831113815,
|
| 330 |
+
"rewards/margins_min": 0.015873130410909653,
|
| 331 |
+
"rewards/margins_std": 0.021959390491247177,
|
| 332 |
+
"rewards/rejected": -0.022160857915878296,
|
| 333 |
+
"step": 170
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.62,
|
| 337 |
+
"grad_norm": 0.306640625,
|
| 338 |
+
"learning_rate": 1.8724253844823456e-07,
|
| 339 |
+
"logits/chosen": -0.10444211959838867,
|
| 340 |
+
"logits/rejected": 0.11585960537195206,
|
| 341 |
+
"logps/chosen": -217.49295043945312,
|
| 342 |
+
"logps/rejected": -236.1336212158203,
|
| 343 |
+
"loss": 0.6789,
|
| 344 |
+
"rewards/accuracies": 0.925000011920929,
|
| 345 |
+
"rewards/chosen": 0.009765163995325565,
|
| 346 |
+
"rewards/margins": 0.02891402505338192,
|
| 347 |
+
"rewards/margins_max": 0.04084194451570511,
|
| 348 |
+
"rewards/margins_min": 0.016986116766929626,
|
| 349 |
+
"rewards/margins_std": 0.01686861552298069,
|
| 350 |
+
"rewards/rejected": -0.019148865714669228,
|
| 351 |
+
"step": 180
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.66,
|
| 355 |
+
"grad_norm": 0.2451171875,
|
| 356 |
+
"learning_rate": 1.5853085673944694e-07,
|
| 357 |
+
"logits/chosen": -0.21767011284828186,
|
| 358 |
+
"logits/rejected": 0.03656459227204323,
|
| 359 |
+
"logps/chosen": -205.89724731445312,
|
| 360 |
+
"logps/rejected": -195.54986572265625,
|
| 361 |
+
"loss": 0.677,
|
| 362 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 363 |
+
"rewards/chosen": 0.009529804810881615,
|
| 364 |
+
"rewards/margins": 0.034868255257606506,
|
| 365 |
+
"rewards/margins_max": 0.05301555246114731,
|
| 366 |
+
"rewards/margins_min": 0.016720956191420555,
|
| 367 |
+
"rewards/margins_std": 0.025664156302809715,
|
| 368 |
+
"rewards/rejected": -0.02533845044672489,
|
| 369 |
+
"step": 190
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.69,
|
| 373 |
+
"grad_norm": 0.2412109375,
|
| 374 |
+
"learning_rate": 1.3115300110997096e-07,
|
| 375 |
+
"logits/chosen": -0.12012083828449249,
|
| 376 |
+
"logits/rejected": 0.10849084705114365,
|
| 377 |
+
"logps/chosen": -219.48818969726562,
|
| 378 |
+
"logps/rejected": -228.03585815429688,
|
| 379 |
+
"loss": 0.6777,
|
| 380 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 381 |
+
"rewards/chosen": 0.007625420577824116,
|
| 382 |
+
"rewards/margins": 0.030659427866339684,
|
| 383 |
+
"rewards/margins_max": 0.044585634022951126,
|
| 384 |
+
"rewards/margins_min": 0.01673322357237339,
|
| 385 |
+
"rewards/margins_std": 0.019694630056619644,
|
| 386 |
+
"rewards/rejected": -0.023034008219838142,
|
| 387 |
+
"step": 200
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.73,
|
| 391 |
+
"grad_norm": 0.2236328125,
|
| 392 |
+
"learning_rate": 1.0550820234444626e-07,
|
| 393 |
+
"logits/chosen": -0.13823586702346802,
|
| 394 |
+
"logits/rejected": 0.07014124095439911,
|
| 395 |
+
"logps/chosen": -209.46163940429688,
|
| 396 |
+
"logps/rejected": -220.68325805664062,
|
| 397 |
+
"loss": 0.6771,
|
| 398 |
+
"rewards/accuracies": 0.949999988079071,
|
| 399 |
+
"rewards/chosen": 0.010470154695212841,
|
| 400 |
+
"rewards/margins": 0.03657643496990204,
|
| 401 |
+
"rewards/margins_max": 0.050303250551223755,
|
| 402 |
+
"rewards/margins_min": 0.02284962125122547,
|
| 403 |
+
"rewards/margins_std": 0.019412647932767868,
|
| 404 |
+
"rewards/rejected": -0.026106279343366623,
|
| 405 |
+
"step": 210
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.76,
|
| 409 |
+
"grad_norm": 0.244140625,
|
| 410 |
+
"learning_rate": 8.197041935593179e-08,
|
| 411 |
+
"logits/chosen": -0.15651056170463562,
|
| 412 |
+
"logits/rejected": 0.06445064395666122,
|
| 413 |
+
"logps/chosen": -212.4806671142578,
|
| 414 |
+
"logps/rejected": -226.1749267578125,
|
| 415 |
+
"loss": 0.6765,
|
| 416 |
+
"rewards/accuracies": 0.925000011920929,
|
| 417 |
+
"rewards/chosen": 0.00890127569437027,
|
| 418 |
+
"rewards/margins": 0.035758793354034424,
|
| 419 |
+
"rewards/margins_max": 0.053012482821941376,
|
| 420 |
+
"rewards/margins_min": 0.01850510574877262,
|
| 421 |
+
"rewards/margins_std": 0.02440040186047554,
|
| 422 |
+
"rewards/rejected": -0.026857519522309303,
|
| 423 |
+
"step": 220
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.8,
|
| 427 |
+
"grad_norm": 0.2333984375,
|
| 428 |
+
"learning_rate": 6.088288602287158e-08,
|
| 429 |
+
"logits/chosen": -0.19696348905563354,
|
| 430 |
+
"logits/rejected": -0.03903040662407875,
|
| 431 |
+
"logps/chosen": -213.6901397705078,
|
| 432 |
+
"logps/rejected": -221.56021118164062,
|
| 433 |
+
"loss": 0.6783,
|
| 434 |
+
"rewards/accuracies": 0.887499988079071,
|
| 435 |
+
"rewards/chosen": 0.009939353913068771,
|
| 436 |
+
"rewards/margins": 0.033906787633895874,
|
| 437 |
+
"rewards/margins_max": 0.05103258416056633,
|
| 438 |
+
"rewards/margins_min": 0.016780991107225418,
|
| 439 |
+
"rewards/margins_std": 0.024219539016485214,
|
| 440 |
+
"rewards/rejected": -0.023967433720827103,
|
| 441 |
+
"step": 230
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.83,
|
| 445 |
+
"grad_norm": 0.2392578125,
|
| 446 |
+
"learning_rate": 4.255310606625123e-08,
|
| 447 |
+
"logits/chosen": -0.18151769042015076,
|
| 448 |
+
"logits/rejected": 0.08627013117074966,
|
| 449 |
+
"logps/chosen": -232.16629028320312,
|
| 450 |
+
"logps/rejected": -224.4107208251953,
|
| 451 |
+
"loss": 0.6781,
|
| 452 |
+
"rewards/accuracies": 0.925000011920929,
|
| 453 |
+
"rewards/chosen": 0.012780356220901012,
|
| 454 |
+
"rewards/margins": 0.03826562687754631,
|
| 455 |
+
"rewards/margins_max": 0.055571459233760834,
|
| 456 |
+
"rewards/margins_min": 0.020959796383976936,
|
| 457 |
+
"rewards/margins_std": 0.024474143981933594,
|
| 458 |
+
"rewards/rejected": -0.025485267862677574,
|
| 459 |
+
"step": 240
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.86,
|
| 463 |
+
"grad_norm": 0.2451171875,
|
| 464 |
+
"learning_rate": 2.724836895290805e-08,
|
| 465 |
+
"logits/chosen": -0.1553465723991394,
|
| 466 |
+
"logits/rejected": 0.18157121539115906,
|
| 467 |
+
"logps/chosen": -247.65811157226562,
|
| 468 |
+
"logps/rejected": -212.87515258789062,
|
| 469 |
+
"loss": 0.6772,
|
| 470 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 471 |
+
"rewards/chosen": 0.011063109152019024,
|
| 472 |
+
"rewards/margins": 0.030740728601813316,
|
| 473 |
+
"rewards/margins_max": 0.044029705226421356,
|
| 474 |
+
"rewards/margins_min": 0.017451755702495575,
|
| 475 |
+
"rewards/margins_std": 0.018793445080518723,
|
| 476 |
+
"rewards/rejected": -0.019677620381116867,
|
| 477 |
+
"step": 250
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.9,
|
| 481 |
+
"grad_norm": 0.2578125,
|
| 482 |
+
"learning_rate": 1.5191852213221196e-08,
|
| 483 |
+
"logits/chosen": -0.13707074522972107,
|
| 484 |
+
"logits/rejected": 0.0773903951048851,
|
| 485 |
+
"logps/chosen": -215.07275390625,
|
| 486 |
+
"logps/rejected": -215.782958984375,
|
| 487 |
+
"loss": 0.6772,
|
| 488 |
+
"rewards/accuracies": 0.862500011920929,
|
| 489 |
+
"rewards/chosen": 0.002797811757773161,
|
| 490 |
+
"rewards/margins": 0.031788431107997894,
|
| 491 |
+
"rewards/margins_max": 0.05227842181921005,
|
| 492 |
+
"rewards/margins_min": 0.01129843108355999,
|
| 493 |
+
"rewards/margins_std": 0.028977233916521072,
|
| 494 |
+
"rewards/rejected": -0.028990617021918297,
|
| 495 |
+
"step": 260
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.93,
|
| 499 |
+
"grad_norm": 0.2373046875,
|
| 500 |
+
"learning_rate": 6.559367010166628e-09,
|
| 501 |
+
"logits/chosen": -0.16215373575687408,
|
| 502 |
+
"logits/rejected": 0.14343757927417755,
|
| 503 |
+
"logps/chosen": -248.9391632080078,
|
| 504 |
+
"logps/rejected": -220.1829833984375,
|
| 505 |
+
"loss": 0.6783,
|
| 506 |
+
"rewards/accuracies": 0.887499988079071,
|
| 507 |
+
"rewards/chosen": 0.009577736258506775,
|
| 508 |
+
"rewards/margins": 0.028629934415221214,
|
| 509 |
+
"rewards/margins_max": 0.045094698667526245,
|
| 510 |
+
"rewards/margins_min": 0.012165175750851631,
|
| 511 |
+
"rewards/margins_std": 0.023284688591957092,
|
| 512 |
+
"rewards/rejected": -0.01905220001935959,
|
| 513 |
+
"step": 270
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.97,
|
| 517 |
+
"grad_norm": 0.2490234375,
|
| 518 |
+
"learning_rate": 1.476794416668703e-09,
|
| 519 |
+
"logits/chosen": -0.1460862010717392,
|
| 520 |
+
"logits/rejected": 0.10489644855260849,
|
| 521 |
+
"logps/chosen": -217.80380249023438,
|
| 522 |
+
"logps/rejected": -210.3385009765625,
|
| 523 |
+
"loss": 0.6764,
|
| 524 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 525 |
+
"rewards/chosen": 0.011174037121236324,
|
| 526 |
+
"rewards/margins": 0.03286902233958244,
|
| 527 |
+
"rewards/margins_max": 0.050892699509859085,
|
| 528 |
+
"rewards/margins_min": 0.014845346100628376,
|
| 529 |
+
"rewards/margins_std": 0.025489334017038345,
|
| 530 |
+
"rewards/rejected": -0.021694988012313843,
|
| 531 |
+
"step": 280
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.0,
|
| 535 |
+
"eval_logits/chosen": 0.7124304175376892,
|
| 536 |
+
"eval_logits/rejected": 0.8419629335403442,
|
| 537 |
+
"eval_logps/chosen": -345.623046875,
|
| 538 |
+
"eval_logps/rejected": -338.8891296386719,
|
| 539 |
+
"eval_loss": 0.6928284764289856,
|
| 540 |
+
"eval_rewards/accuracies": 0.5158730149269104,
|
| 541 |
+
"eval_rewards/chosen": -0.00016850981046445668,
|
| 542 |
+
"eval_rewards/margins": -0.00010753136302810162,
|
| 543 |
+
"eval_rewards/margins_max": 0.019001232460141182,
|
| 544 |
+
"eval_rewards/margins_min": -0.01899593509733677,
|
| 545 |
+
"eval_rewards/margins_std": 0.012460649013519287,
|
| 546 |
+
"eval_rewards/rejected": -6.097855293774046e-05,
|
| 547 |
+
"eval_runtime": 224.8356,
|
| 548 |
+
"eval_samples_per_second": 17.791,
|
| 549 |
+
"eval_steps_per_second": 0.28,
|
| 550 |
+
"step": 289
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.0,
|
| 554 |
+
"step": 289,
|
| 555 |
+
"total_flos": 0.0,
|
| 556 |
+
"train_loss": 0.6824996727148142,
|
| 557 |
+
"train_runtime": 3006.8067,
|
| 558 |
+
"train_samples_per_second": 6.156,
|
| 559 |
+
"train_steps_per_second": 0.096
|
| 560 |
+
}
|
| 561 |
+
],
|
| 562 |
+
"logging_steps": 10,
|
| 563 |
+
"max_steps": 289,
|
| 564 |
+
"num_input_tokens_seen": 0,
|
| 565 |
+
"num_train_epochs": 1,
|
| 566 |
+
"save_steps": 100,
|
| 567 |
+
"total_flos": 0.0,
|
| 568 |
+
"train_batch_size": 2,
|
| 569 |
+
"trial_name": null,
|
| 570 |
+
"trial_params": null
|
| 571 |
+
}
|