Model save
Browse files- README.md +61 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- runs/May18_00-29-15_gpu4-119-5/events.out.tfevents.1715956269.gpu4-119-5.1736445.0 +2 -2
- train_results.json +8 -0
- trainer_state.json +618 -0
README.md
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
tags:
|
| 4 |
+
- trl
|
| 5 |
+
- dpo
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
base_model: DUAL-GPO-2/phi-2-gpo-v34-merged-i1
|
| 8 |
+
model-index:
|
| 9 |
+
- name: phi-2-gpo-v5-i2
|
| 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 |
+
# phi-2-gpo-v5-i2
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [DUAL-GPO-2/phi-2-gpo-v34-merged-i1](https://huggingface.co/DUAL-GPO-2/phi-2-gpo-v34-merged-i1) on the None 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: 5e-06
|
| 38 |
+
- train_batch_size: 4
|
| 39 |
+
- eval_batch_size: 4
|
| 40 |
+
- seed: 42
|
| 41 |
+
- distributed_type: multi-GPU
|
| 42 |
+
- num_devices: 3
|
| 43 |
+
- gradient_accumulation_steps: 4
|
| 44 |
+
- total_train_batch_size: 48
|
| 45 |
+
- total_eval_batch_size: 12
|
| 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: 1
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
### Framework versions
|
| 56 |
+
|
| 57 |
+
- PEFT 0.7.1
|
| 58 |
+
- Transformers 4.36.2
|
| 59 |
+
- Pytorch 2.1.2+cu121
|
| 60 |
+
- Datasets 2.14.6
|
| 61 |
+
- 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 167807296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7bca136dec62bd9ea71a8306cfbc3a40f7a8f9f9df1d3c0e161798b17ba5c7e
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.13873135742659753,
|
| 4 |
+
"train_runtime": 5562.1192,
|
| 5 |
+
"train_samples": 20000,
|
| 6 |
+
"train_samples_per_second": 3.596,
|
| 7 |
+
"train_steps_per_second": 0.075
|
| 8 |
+
}
|
runs/May18_00-29-15_gpu4-119-5/events.out.tfevents.1715956269.gpu4-119-5.1736445.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:160fdfbdb807c7d70cadcaba72486ffd7886ee3892b34f58d046556551fe6a7e
|
| 3 |
+
size 31679
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.13873135742659753,
|
| 4 |
+
"train_runtime": 5562.1192,
|
| 5 |
+
"train_samples": 20000,
|
| 6 |
+
"train_samples_per_second": 3.596,
|
| 7 |
+
"train_steps_per_second": 0.075
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,618 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9982003599280144,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 416,
|
| 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 |
+
"learning_rate": 1.1904761904761906e-07,
|
| 14 |
+
"logits/chosen": -0.10997498780488968,
|
| 15 |
+
"logits/rejected": -0.14968934655189514,
|
| 16 |
+
"logps/chosen": -470.78253173828125,
|
| 17 |
+
"logps/rejected": -601.5414428710938,
|
| 18 |
+
"loss": 0.21,
|
| 19 |
+
"rewards/accuracies": 0.0,
|
| 20 |
+
"rewards/chosen": 0.0,
|
| 21 |
+
"rewards/margins": 0.0,
|
| 22 |
+
"rewards/rejected": 0.0,
|
| 23 |
+
"step": 1
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02,
|
| 27 |
+
"learning_rate": 1.1904761904761906e-06,
|
| 28 |
+
"logits/chosen": -0.1849229484796524,
|
| 29 |
+
"logits/rejected": -0.13570153713226318,
|
| 30 |
+
"logps/chosen": -495.9288330078125,
|
| 31 |
+
"logps/rejected": -719.5632934570312,
|
| 32 |
+
"loss": 0.2073,
|
| 33 |
+
"rewards/accuracies": 0.4027777910232544,
|
| 34 |
+
"rewards/chosen": -1.4412669315788662e-06,
|
| 35 |
+
"rewards/margins": -4.9562087951926515e-05,
|
| 36 |
+
"rewards/rejected": 4.812082988792099e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05,
|
| 41 |
+
"learning_rate": 2.380952380952381e-06,
|
| 42 |
+
"logits/chosen": -0.21864363551139832,
|
| 43 |
+
"logits/rejected": -0.1575668752193451,
|
| 44 |
+
"logps/chosen": -500.5243225097656,
|
| 45 |
+
"logps/rejected": -724.0389404296875,
|
| 46 |
+
"loss": 0.2106,
|
| 47 |
+
"rewards/accuracies": 0.65625,
|
| 48 |
+
"rewards/chosen": -0.0024510668590664864,
|
| 49 |
+
"rewards/margins": 0.002445765770971775,
|
| 50 |
+
"rewards/rejected": -0.004896833095699549,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07,
|
| 55 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 56 |
+
"logits/chosen": -0.11167863756418228,
|
| 57 |
+
"logits/rejected": -0.1275203675031662,
|
| 58 |
+
"logps/chosen": -545.2283935546875,
|
| 59 |
+
"logps/rejected": -782.0553588867188,
|
| 60 |
+
"loss": 0.2009,
|
| 61 |
+
"rewards/accuracies": 0.731249988079071,
|
| 62 |
+
"rewards/chosen": -0.013457772321999073,
|
| 63 |
+
"rewards/margins": 0.01753374934196472,
|
| 64 |
+
"rewards/rejected": -0.03099152073264122,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1,
|
| 69 |
+
"learning_rate": 4.761904761904762e-06,
|
| 70 |
+
"logits/chosen": -0.20949259400367737,
|
| 71 |
+
"logits/rejected": -0.17874333262443542,
|
| 72 |
+
"logps/chosen": -546.1171875,
|
| 73 |
+
"logps/rejected": -841.0467529296875,
|
| 74 |
+
"loss": 0.1651,
|
| 75 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 76 |
+
"rewards/chosen": -0.05306588485836983,
|
| 77 |
+
"rewards/margins": 0.06392969191074371,
|
| 78 |
+
"rewards/rejected": -0.11699558794498444,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12,
|
| 83 |
+
"learning_rate": 4.994357350311441e-06,
|
| 84 |
+
"logits/chosen": -0.19548015296459198,
|
| 85 |
+
"logits/rejected": -0.24876411259174347,
|
| 86 |
+
"logps/chosen": -626.0526123046875,
|
| 87 |
+
"logps/rejected": -1041.28076171875,
|
| 88 |
+
"loss": 0.1523,
|
| 89 |
+
"rewards/accuracies": 0.731249988079071,
|
| 90 |
+
"rewards/chosen": -0.1381954401731491,
|
| 91 |
+
"rewards/margins": 0.14297422766685486,
|
| 92 |
+
"rewards/rejected": -0.28116968274116516,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.14,
|
| 97 |
+
"learning_rate": 4.97147773390341e-06,
|
| 98 |
+
"logits/chosen": -0.2218095362186432,
|
| 99 |
+
"logits/rejected": -0.2248532772064209,
|
| 100 |
+
"logps/chosen": -627.7503662109375,
|
| 101 |
+
"logps/rejected": -929.7164306640625,
|
| 102 |
+
"loss": 0.1461,
|
| 103 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 104 |
+
"rewards/chosen": -0.14269283413887024,
|
| 105 |
+
"rewards/margins": 0.11241813004016876,
|
| 106 |
+
"rewards/rejected": -0.2551110088825226,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.17,
|
| 111 |
+
"learning_rate": 4.931169703639282e-06,
|
| 112 |
+
"logits/chosen": -0.23902878165245056,
|
| 113 |
+
"logits/rejected": -0.19728106260299683,
|
| 114 |
+
"logps/chosen": -686.7578735351562,
|
| 115 |
+
"logps/rejected": -1015.8434448242188,
|
| 116 |
+
"loss": 0.1503,
|
| 117 |
+
"rewards/accuracies": 0.737500011920929,
|
| 118 |
+
"rewards/chosen": -0.1398530751466751,
|
| 119 |
+
"rewards/margins": 0.1030840128660202,
|
| 120 |
+
"rewards/rejected": -0.24293705821037292,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.19,
|
| 125 |
+
"learning_rate": 4.873717504456219e-06,
|
| 126 |
+
"logits/chosen": -0.22387316823005676,
|
| 127 |
+
"logits/rejected": -0.2015857696533203,
|
| 128 |
+
"logps/chosen": -625.0352783203125,
|
| 129 |
+
"logps/rejected": -1033.6270751953125,
|
| 130 |
+
"loss": 0.1483,
|
| 131 |
+
"rewards/accuracies": 0.768750011920929,
|
| 132 |
+
"rewards/chosen": -0.1254601627588272,
|
| 133 |
+
"rewards/margins": 0.12839995324611664,
|
| 134 |
+
"rewards/rejected": -0.25386008620262146,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.22,
|
| 139 |
+
"learning_rate": 4.7995262788689865e-06,
|
| 140 |
+
"logits/chosen": -0.27522459626197815,
|
| 141 |
+
"logits/rejected": -0.21160630881786346,
|
| 142 |
+
"logps/chosen": -697.7965698242188,
|
| 143 |
+
"logps/rejected": -1014.5755004882812,
|
| 144 |
+
"loss": 0.1439,
|
| 145 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 146 |
+
"rewards/chosen": -0.1593645215034485,
|
| 147 |
+
"rewards/margins": 0.13819821178913116,
|
| 148 |
+
"rewards/rejected": -0.29756274819374084,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.24,
|
| 153 |
+
"learning_rate": 4.709119209978242e-06,
|
| 154 |
+
"logits/chosen": -0.25298193097114563,
|
| 155 |
+
"logits/rejected": -0.2807552218437195,
|
| 156 |
+
"logps/chosen": -657.3956298828125,
|
| 157 |
+
"logps/rejected": -1090.194091796875,
|
| 158 |
+
"loss": 0.1371,
|
| 159 |
+
"rewards/accuracies": 0.768750011920929,
|
| 160 |
+
"rewards/chosen": -0.15086904168128967,
|
| 161 |
+
"rewards/margins": 0.16771209239959717,
|
| 162 |
+
"rewards/rejected": -0.31858113408088684,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.26,
|
| 167 |
+
"learning_rate": 4.603133832077953e-06,
|
| 168 |
+
"logits/chosen": -0.2874082922935486,
|
| 169 |
+
"logits/rejected": -0.18505045771598816,
|
| 170 |
+
"logps/chosen": -685.4857788085938,
|
| 171 |
+
"logps/rejected": -932.3782958984375,
|
| 172 |
+
"loss": 0.1492,
|
| 173 |
+
"rewards/accuracies": 0.675000011920929,
|
| 174 |
+
"rewards/chosen": -0.1552571952342987,
|
| 175 |
+
"rewards/margins": 0.113719142973423,
|
| 176 |
+
"rewards/rejected": -0.2689763307571411,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.29,
|
| 181 |
+
"learning_rate": 4.482317534878901e-06,
|
| 182 |
+
"logits/chosen": -0.2046702355146408,
|
| 183 |
+
"logits/rejected": -0.21371881663799286,
|
| 184 |
+
"logps/chosen": -615.016357421875,
|
| 185 |
+
"logps/rejected": -1014.6843872070312,
|
| 186 |
+
"loss": 0.1446,
|
| 187 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 188 |
+
"rewards/chosen": -0.13098101317882538,
|
| 189 |
+
"rewards/margins": 0.15675313770771027,
|
| 190 |
+
"rewards/rejected": -0.2877342104911804,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.31,
|
| 195 |
+
"learning_rate": 4.3475222930516484e-06,
|
| 196 |
+
"logits/chosen": -0.30239051580429077,
|
| 197 |
+
"logits/rejected": -0.2552723288536072,
|
| 198 |
+
"logps/chosen": -661.1831665039062,
|
| 199 |
+
"logps/rejected": -1063.7322998046875,
|
| 200 |
+
"loss": 0.1373,
|
| 201 |
+
"rewards/accuracies": 0.78125,
|
| 202 |
+
"rewards/chosen": -0.16380271315574646,
|
| 203 |
+
"rewards/margins": 0.14136558771133423,
|
| 204 |
+
"rewards/rejected": -0.3051683008670807,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.34,
|
| 209 |
+
"learning_rate": 4.199698658255298e-06,
|
| 210 |
+
"logits/chosen": -0.23284006118774414,
|
| 211 |
+
"logits/rejected": -0.24393205344676971,
|
| 212 |
+
"logps/chosen": -689.3310546875,
|
| 213 |
+
"logps/rejected": -1185.5355224609375,
|
| 214 |
+
"loss": 0.1333,
|
| 215 |
+
"rewards/accuracies": 0.8062499761581421,
|
| 216 |
+
"rewards/chosen": -0.1679478883743286,
|
| 217 |
+
"rewards/margins": 0.18742243945598602,
|
| 218 |
+
"rewards/rejected": -0.35537031292915344,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.36,
|
| 223 |
+
"learning_rate": 4.039889056019159e-06,
|
| 224 |
+
"logits/chosen": -0.20814350247383118,
|
| 225 |
+
"logits/rejected": -0.24712209403514862,
|
| 226 |
+
"logps/chosen": -631.2597045898438,
|
| 227 |
+
"logps/rejected": -1041.0517578125,
|
| 228 |
+
"loss": 0.1331,
|
| 229 |
+
"rewards/accuracies": 0.793749988079071,
|
| 230 |
+
"rewards/chosen": -0.15335983037948608,
|
| 231 |
+
"rewards/margins": 0.17034170031547546,
|
| 232 |
+
"rewards/rejected": -0.32370156049728394,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.38,
|
| 237 |
+
"learning_rate": 3.869220434746509e-06,
|
| 238 |
+
"logits/chosen": -0.2555493712425232,
|
| 239 |
+
"logits/rejected": -0.2478325366973877,
|
| 240 |
+
"logps/chosen": -709.0995483398438,
|
| 241 |
+
"logps/rejected": -1125.80615234375,
|
| 242 |
+
"loss": 0.1273,
|
| 243 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 244 |
+
"rewards/chosen": -0.16779521107673645,
|
| 245 |
+
"rewards/margins": 0.157444566488266,
|
| 246 |
+
"rewards/rejected": -0.32523980736732483,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.41,
|
| 251 |
+
"learning_rate": 3.688896318678322e-06,
|
| 252 |
+
"logits/chosen": -0.24241690337657928,
|
| 253 |
+
"logits/rejected": -0.2430168092250824,
|
| 254 |
+
"logps/chosen": -652.2470703125,
|
| 255 |
+
"logps/rejected": -1029.941162109375,
|
| 256 |
+
"loss": 0.1405,
|
| 257 |
+
"rewards/accuracies": 0.762499988079071,
|
| 258 |
+
"rewards/chosen": -0.1708507239818573,
|
| 259 |
+
"rewards/margins": 0.14247409999370575,
|
| 260 |
+
"rewards/rejected": -0.31332483887672424,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.43,
|
| 265 |
+
"learning_rate": 3.5001883208580668e-06,
|
| 266 |
+
"logits/chosen": -0.2642208933830261,
|
| 267 |
+
"logits/rejected": -0.2481091320514679,
|
| 268 |
+
"logps/chosen": -667.6624145507812,
|
| 269 |
+
"logps/rejected": -1021.7717895507812,
|
| 270 |
+
"loss": 0.1306,
|
| 271 |
+
"rewards/accuracies": 0.762499988079071,
|
| 272 |
+
"rewards/chosen": -0.18226784467697144,
|
| 273 |
+
"rewards/margins": 0.14150500297546387,
|
| 274 |
+
"rewards/rejected": -0.3237728476524353,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.46,
|
| 279 |
+
"learning_rate": 3.30442717594657e-06,
|
| 280 |
+
"logits/chosen": -0.1869751662015915,
|
| 281 |
+
"logits/rejected": -0.23872609436511993,
|
| 282 |
+
"logps/chosen": -644.8170776367188,
|
| 283 |
+
"logps/rejected": -1029.4417724609375,
|
| 284 |
+
"loss": 0.1211,
|
| 285 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 286 |
+
"rewards/chosen": -0.1459895819425583,
|
| 287 |
+
"rewards/margins": 0.15111900866031647,
|
| 288 |
+
"rewards/rejected": -0.29710859060287476,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.48,
|
| 293 |
+
"learning_rate": 3.102993356121938e-06,
|
| 294 |
+
"logits/chosen": -0.260442852973938,
|
| 295 |
+
"logits/rejected": -0.2213623970746994,
|
| 296 |
+
"logps/chosen": -672.3732299804688,
|
| 297 |
+
"logps/rejected": -1037.558837890625,
|
| 298 |
+
"loss": 0.1341,
|
| 299 |
+
"rewards/accuracies": 0.793749988079071,
|
| 300 |
+
"rewards/chosen": -0.1695290207862854,
|
| 301 |
+
"rewards/margins": 0.14989891648292542,
|
| 302 |
+
"rewards/rejected": -0.3194279074668884,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5,
|
| 307 |
+
"learning_rate": 2.8973073362395e-06,
|
| 308 |
+
"logits/chosen": -0.25881263613700867,
|
| 309 |
+
"logits/rejected": -0.20856423676013947,
|
| 310 |
+
"logps/chosen": -658.4823608398438,
|
| 311 |
+
"logps/rejected": -1057.578369140625,
|
| 312 |
+
"loss": 0.1328,
|
| 313 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 314 |
+
"rewards/chosen": -0.17251142859458923,
|
| 315 |
+
"rewards/margins": 0.17602583765983582,
|
| 316 |
+
"rewards/rejected": -0.34853726625442505,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.53,
|
| 321 |
+
"learning_rate": 2.6888195769001147e-06,
|
| 322 |
+
"logits/chosen": -0.2791309952735901,
|
| 323 |
+
"logits/rejected": -0.27674758434295654,
|
| 324 |
+
"logps/chosen": -642.2847900390625,
|
| 325 |
+
"logps/rejected": -1090.7037353515625,
|
| 326 |
+
"loss": 0.1269,
|
| 327 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 328 |
+
"rewards/chosen": -0.18547888100147247,
|
| 329 |
+
"rewards/margins": 0.1763353794813156,
|
| 330 |
+
"rewards/rejected": -0.3618142306804657,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.55,
|
| 335 |
+
"learning_rate": 2.479000296064417e-06,
|
| 336 |
+
"logits/chosen": -0.24742920696735382,
|
| 337 |
+
"logits/rejected": -0.2712458372116089,
|
| 338 |
+
"logps/chosen": -732.4454956054688,
|
| 339 |
+
"logps/rejected": -1207.4635009765625,
|
| 340 |
+
"loss": 0.1238,
|
| 341 |
+
"rewards/accuracies": 0.856249988079071,
|
| 342 |
+
"rewards/chosen": -0.18814927339553833,
|
| 343 |
+
"rewards/margins": 0.19270290434360504,
|
| 344 |
+
"rewards/rejected": -0.38085219264030457,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.58,
|
| 349 |
+
"learning_rate": 2.269329101341745e-06,
|
| 350 |
+
"logits/chosen": -0.30566203594207764,
|
| 351 |
+
"logits/rejected": -0.2690572142601013,
|
| 352 |
+
"logps/chosen": -696.6936645507812,
|
| 353 |
+
"logps/rejected": -1079.8812255859375,
|
| 354 |
+
"loss": 0.1214,
|
| 355 |
+
"rewards/accuracies": 0.824999988079071,
|
| 356 |
+
"rewards/chosen": -0.196825310587883,
|
| 357 |
+
"rewards/margins": 0.17798085510730743,
|
| 358 |
+
"rewards/rejected": -0.37480616569519043,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6,
|
| 363 |
+
"learning_rate": 2.06128455606496e-06,
|
| 364 |
+
"logits/chosen": -0.23360852897167206,
|
| 365 |
+
"logits/rejected": -0.20390859246253967,
|
| 366 |
+
"logps/chosen": -674.8441162109375,
|
| 367 |
+
"logps/rejected": -1029.4664306640625,
|
| 368 |
+
"loss": 0.1324,
|
| 369 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 370 |
+
"rewards/chosen": -0.1800800859928131,
|
| 371 |
+
"rewards/margins": 0.17153646051883698,
|
| 372 |
+
"rewards/rejected": -0.3516165018081665,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.62,
|
| 377 |
+
"learning_rate": 1.856333752729311e-06,
|
| 378 |
+
"logits/chosen": -0.23972924053668976,
|
| 379 |
+
"logits/rejected": -0.2775985598564148,
|
| 380 |
+
"logps/chosen": -682.693115234375,
|
| 381 |
+
"logps/rejected": -1193.5947265625,
|
| 382 |
+
"loss": 0.1254,
|
| 383 |
+
"rewards/accuracies": 0.800000011920929,
|
| 384 |
+
"rewards/chosen": -0.19959069788455963,
|
| 385 |
+
"rewards/margins": 0.19032658636569977,
|
| 386 |
+
"rewards/rejected": -0.3899172842502594,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.65,
|
| 391 |
+
"learning_rate": 1.6559219673215784e-06,
|
| 392 |
+
"logits/chosen": -0.30556970834732056,
|
| 393 |
+
"logits/rejected": -0.3009134829044342,
|
| 394 |
+
"logps/chosen": -681.2161865234375,
|
| 395 |
+
"logps/rejected": -1142.70361328125,
|
| 396 |
+
"loss": 0.1318,
|
| 397 |
+
"rewards/accuracies": 0.793749988079071,
|
| 398 |
+
"rewards/chosen": -0.20223207771778107,
|
| 399 |
+
"rewards/margins": 0.17814533412456512,
|
| 400 |
+
"rewards/rejected": -0.3803774416446686,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.67,
|
| 405 |
+
"learning_rate": 1.4614624674952843e-06,
|
| 406 |
+
"logits/chosen": -0.26370662450790405,
|
| 407 |
+
"logits/rejected": -0.23301748931407928,
|
| 408 |
+
"logps/chosen": -725.9805908203125,
|
| 409 |
+
"logps/rejected": -1188.698974609375,
|
| 410 |
+
"loss": 0.1291,
|
| 411 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 412 |
+
"rewards/chosen": -0.20386967062950134,
|
| 413 |
+
"rewards/margins": 0.1635088473558426,
|
| 414 |
+
"rewards/rejected": -0.3673785328865051,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7,
|
| 419 |
+
"learning_rate": 1.2743265464628787e-06,
|
| 420 |
+
"logits/chosen": -0.26803213357925415,
|
| 421 |
+
"logits/rejected": -0.23297333717346191,
|
| 422 |
+
"logps/chosen": -660.8951416015625,
|
| 423 |
+
"logps/rejected": -1116.446533203125,
|
| 424 |
+
"loss": 0.1294,
|
| 425 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 426 |
+
"rewards/chosen": -0.1704072505235672,
|
| 427 |
+
"rewards/margins": 0.18539056181907654,
|
| 428 |
+
"rewards/rejected": -0.35579782724380493,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.72,
|
| 433 |
+
"learning_rate": 1.0958338528840893e-06,
|
| 434 |
+
"logits/chosen": -0.23666846752166748,
|
| 435 |
+
"logits/rejected": -0.264843612909317,
|
| 436 |
+
"logps/chosen": -715.3922119140625,
|
| 437 |
+
"logps/rejected": -1168.862548828125,
|
| 438 |
+
"loss": 0.1267,
|
| 439 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 440 |
+
"rewards/chosen": -0.18868893384933472,
|
| 441 |
+
"rewards/margins": 0.19361743330955505,
|
| 442 |
+
"rewards/rejected": -0.38230639696121216,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.74,
|
| 447 |
+
"learning_rate": 9.272430849423175e-07,
|
| 448 |
+
"logits/chosen": -0.2275594025850296,
|
| 449 |
+
"logits/rejected": -0.20153971016407013,
|
| 450 |
+
"logps/chosen": -691.9971313476562,
|
| 451 |
+
"logps/rejected": -1152.3519287109375,
|
| 452 |
+
"loss": 0.1327,
|
| 453 |
+
"rewards/accuracies": 0.768750011920929,
|
| 454 |
+
"rewards/chosen": -0.1932949721813202,
|
| 455 |
+
"rewards/margins": 0.19385525584220886,
|
| 456 |
+
"rewards/rejected": -0.38715022802352905,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.77,
|
| 461 |
+
"learning_rate": 7.697431142327633e-07,
|
| 462 |
+
"logits/chosen": -0.2559944987297058,
|
| 463 |
+
"logits/rejected": -0.2104511708021164,
|
| 464 |
+
"logps/chosen": -690.9197998046875,
|
| 465 |
+
"logps/rejected": -1113.470703125,
|
| 466 |
+
"loss": 0.1352,
|
| 467 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 468 |
+
"rewards/chosen": -0.19007542729377747,
|
| 469 |
+
"rewards/margins": 0.16325220465660095,
|
| 470 |
+
"rewards/rejected": -0.3533276319503784,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.79,
|
| 475 |
+
"learning_rate": 6.244446020550182e-07,
|
| 476 |
+
"logits/chosen": -0.21340703964233398,
|
| 477 |
+
"logits/rejected": -0.23157565295696259,
|
| 478 |
+
"logps/chosen": -706.4766235351562,
|
| 479 |
+
"logps/rejected": -1090.2486572265625,
|
| 480 |
+
"loss": 0.1214,
|
| 481 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 482 |
+
"rewards/chosen": -0.19076435267925262,
|
| 483 |
+
"rewards/margins": 0.16976633667945862,
|
| 484 |
+
"rewards/rejected": -0.36053070425987244,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.82,
|
| 489 |
+
"learning_rate": 4.923721672305148e-07,
|
| 490 |
+
"logits/chosen": -0.2297631800174713,
|
| 491 |
+
"logits/rejected": -0.18899887800216675,
|
| 492 |
+
"logps/chosen": -723.7625732421875,
|
| 493 |
+
"logps/rejected": -1124.594482421875,
|
| 494 |
+
"loss": 0.1285,
|
| 495 |
+
"rewards/accuracies": 0.768750011920929,
|
| 496 |
+
"rewards/chosen": -0.19292166829109192,
|
| 497 |
+
"rewards/margins": 0.18659308552742004,
|
| 498 |
+
"rewards/rejected": -0.37951475381851196,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.84,
|
| 503 |
+
"learning_rate": 3.7445716067596506e-07,
|
| 504 |
+
"logits/chosen": -0.22677800059318542,
|
| 505 |
+
"logits/rejected": -0.22308814525604248,
|
| 506 |
+
"logps/chosen": -663.9019775390625,
|
| 507 |
+
"logps/rejected": -1152.65283203125,
|
| 508 |
+
"loss": 0.1117,
|
| 509 |
+
"rewards/accuracies": 0.84375,
|
| 510 |
+
"rewards/chosen": -0.1815580129623413,
|
| 511 |
+
"rewards/margins": 0.20128455758094788,
|
| 512 |
+
"rewards/rejected": -0.3828425407409668,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.86,
|
| 517 |
+
"learning_rate": 2.7153109768518926e-07,
|
| 518 |
+
"logits/chosen": -0.249908447265625,
|
| 519 |
+
"logits/rejected": -0.1893308460712433,
|
| 520 |
+
"logps/chosen": -674.8004150390625,
|
| 521 |
+
"logps/rejected": -1103.431396484375,
|
| 522 |
+
"loss": 0.1331,
|
| 523 |
+
"rewards/accuracies": 0.768750011920929,
|
| 524 |
+
"rewards/chosen": -0.18865013122558594,
|
| 525 |
+
"rewards/margins": 0.16974930465221405,
|
| 526 |
+
"rewards/rejected": -0.3583994507789612,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.89,
|
| 531 |
+
"learning_rate": 1.8431979423369607e-07,
|
| 532 |
+
"logits/chosen": -0.20425882935523987,
|
| 533 |
+
"logits/rejected": -0.2509860098361969,
|
| 534 |
+
"logps/chosen": -787.9830322265625,
|
| 535 |
+
"logps/rejected": -1210.3958740234375,
|
| 536 |
+
"loss": 0.1263,
|
| 537 |
+
"rewards/accuracies": 0.831250011920929,
|
| 538 |
+
"rewards/chosen": -0.20410247147083282,
|
| 539 |
+
"rewards/margins": 0.1842099130153656,
|
| 540 |
+
"rewards/rejected": -0.3883123993873596,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.91,
|
| 545 |
+
"learning_rate": 1.1343824865573422e-07,
|
| 546 |
+
"logits/chosen": -0.23611795902252197,
|
| 547 |
+
"logits/rejected": -0.2374495565891266,
|
| 548 |
+
"logps/chosen": -683.4967041015625,
|
| 549 |
+
"logps/rejected": -1033.9107666015625,
|
| 550 |
+
"loss": 0.1211,
|
| 551 |
+
"rewards/accuracies": 0.762499988079071,
|
| 552 |
+
"rewards/chosen": -0.18874987959861755,
|
| 553 |
+
"rewards/margins": 0.15766242146492004,
|
| 554 |
+
"rewards/rejected": -0.34641233086586,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.94,
|
| 559 |
+
"learning_rate": 5.9386304787299175e-08,
|
| 560 |
+
"logits/chosen": -0.2402607947587967,
|
| 561 |
+
"logits/rejected": -0.2553133964538574,
|
| 562 |
+
"logps/chosen": -653.82177734375,
|
| 563 |
+
"logps/rejected": -1058.991943359375,
|
| 564 |
+
"loss": 0.1348,
|
| 565 |
+
"rewards/accuracies": 0.762499988079071,
|
| 566 |
+
"rewards/chosen": -0.18406951427459717,
|
| 567 |
+
"rewards/margins": 0.15722426772117615,
|
| 568 |
+
"rewards/rejected": -0.3412937819957733,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.96,
|
| 573 |
+
"learning_rate": 2.2545127157831416e-08,
|
| 574 |
+
"logits/chosen": -0.2702855169773102,
|
| 575 |
+
"logits/rejected": -0.2571588158607483,
|
| 576 |
+
"logps/chosen": -723.2869262695312,
|
| 577 |
+
"logps/rejected": -1087.939208984375,
|
| 578 |
+
"loss": 0.1216,
|
| 579 |
+
"rewards/accuracies": 0.762499988079071,
|
| 580 |
+
"rewards/chosen": -0.19856762886047363,
|
| 581 |
+
"rewards/margins": 0.1647205799818039,
|
| 582 |
+
"rewards/rejected": -0.36328819394111633,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.98,
|
| 587 |
+
"learning_rate": 3.1745130869123564e-09,
|
| 588 |
+
"logits/chosen": -0.2467096745967865,
|
| 589 |
+
"logits/rejected": -0.2630787193775177,
|
| 590 |
+
"logps/chosen": -715.5072021484375,
|
| 591 |
+
"logps/rejected": -1024.0079345703125,
|
| 592 |
+
"loss": 0.1273,
|
| 593 |
+
"rewards/accuracies": 0.75,
|
| 594 |
+
"rewards/chosen": -0.20651917159557343,
|
| 595 |
+
"rewards/margins": 0.14182263612747192,
|
| 596 |
+
"rewards/rejected": -0.34834182262420654,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.0,
|
| 601 |
+
"step": 416,
|
| 602 |
+
"total_flos": 0.0,
|
| 603 |
+
"train_loss": 0.13873135742659753,
|
| 604 |
+
"train_runtime": 5562.1192,
|
| 605 |
+
"train_samples_per_second": 3.596,
|
| 606 |
+
"train_steps_per_second": 0.075
|
| 607 |
+
}
|
| 608 |
+
],
|
| 609 |
+
"logging_steps": 10,
|
| 610 |
+
"max_steps": 416,
|
| 611 |
+
"num_input_tokens_seen": 0,
|
| 612 |
+
"num_train_epochs": 1,
|
| 613 |
+
"save_steps": 100,
|
| 614 |
+
"total_flos": 0.0,
|
| 615 |
+
"train_batch_size": 4,
|
| 616 |
+
"trial_name": null,
|
| 617 |
+
"trial_params": null
|
| 618 |
+
}
|