Model save
Browse files- README.md +62 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- runs/May16_14-53-02_gpu4-119-5/events.out.tfevents.1715835303.gpu4-119-5.182802.0 +2 -2
- train_results.json +8 -0
- trainer_state.json +618 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: DUAL-GPO/phi-2-gpo-new-i0
|
| 9 |
+
model-index:
|
| 10 |
+
- name: phi-2-gpo-v33-i1
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# phi-2-gpo-v33-i1
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [DUAL-GPO/phi-2-gpo-new-i0](https://huggingface.co/DUAL-GPO/phi-2-gpo-new-i0) on the None dataset.
|
| 20 |
+
|
| 21 |
+
## Model description
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Intended uses & limitations
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training and evaluation data
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training procedure
|
| 34 |
+
|
| 35 |
+
### Training hyperparameters
|
| 36 |
+
|
| 37 |
+
The following hyperparameters were used during training:
|
| 38 |
+
- learning_rate: 5e-06
|
| 39 |
+
- train_batch_size: 4
|
| 40 |
+
- eval_batch_size: 4
|
| 41 |
+
- seed: 42
|
| 42 |
+
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 3
|
| 44 |
+
- gradient_accumulation_steps: 4
|
| 45 |
+
- total_train_batch_size: 48
|
| 46 |
+
- total_eval_batch_size: 12
|
| 47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
+
- lr_scheduler_type: cosine
|
| 49 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 50 |
+
- num_epochs: 1
|
| 51 |
+
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
### Framework versions
|
| 57 |
+
|
| 58 |
+
- PEFT 0.7.1
|
| 59 |
+
- Transformers 4.36.2
|
| 60 |
+
- Pytorch 2.1.2+cu121
|
| 61 |
+
- Datasets 2.14.6
|
| 62 |
+
- 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 sha256:3209a18dea76cae78db4c3ebfcf7fb62bd90d85f0c7f148a28183de3afe41046
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.3107354554276054,
|
| 4 |
+
"train_runtime": 5919.6286,
|
| 5 |
+
"train_samples": 20000,
|
| 6 |
+
"train_samples_per_second": 3.379,
|
| 7 |
+
"train_steps_per_second": 0.07
|
| 8 |
+
}
|
runs/May16_14-53-02_gpu4-119-5/events.out.tfevents.1715835303.gpu4-119-5.182802.0
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 sha256:797c7e887d3df16805ab8d0c7a1547e507977242ffca88a4388c1108b59b87ad
|
| 3 |
+
size 31674
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.3107354554276054,
|
| 4 |
+
"train_runtime": 5919.6286,
|
| 5 |
+
"train_samples": 20000,
|
| 6 |
+
"train_samples_per_second": 3.379,
|
| 7 |
+
"train_steps_per_second": 0.07
|
| 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": 0.06842132657766342,
|
| 15 |
+
"logits/rejected": 0.05148967728018761,
|
| 16 |
+
"logps/chosen": -254.1962432861328,
|
| 17 |
+
"logps/rejected": -268.0105285644531,
|
| 18 |
+
"loss": 0.3778,
|
| 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": 0.14202657341957092,
|
| 29 |
+
"logits/rejected": 0.2216137945652008,
|
| 30 |
+
"logps/chosen": -380.2115478515625,
|
| 31 |
+
"logps/rejected": -331.30743408203125,
|
| 32 |
+
"loss": 0.3702,
|
| 33 |
+
"rewards/accuracies": 0.3819444477558136,
|
| 34 |
+
"rewards/chosen": 5.156885163160041e-05,
|
| 35 |
+
"rewards/margins": -8.286008232971653e-05,
|
| 36 |
+
"rewards/rejected": 0.00013442893396131694,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05,
|
| 41 |
+
"learning_rate": 2.380952380952381e-06,
|
| 42 |
+
"logits/chosen": 0.11058167368173599,
|
| 43 |
+
"logits/rejected": 0.1396401971578598,
|
| 44 |
+
"logps/chosen": -336.24676513671875,
|
| 45 |
+
"logps/rejected": -305.8110656738281,
|
| 46 |
+
"loss": 0.3688,
|
| 47 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 48 |
+
"rewards/chosen": 2.6639914722181857e-05,
|
| 49 |
+
"rewards/margins": 8.525094017386436e-07,
|
| 50 |
+
"rewards/rejected": 2.578740895842202e-05,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07,
|
| 55 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 56 |
+
"logits/chosen": 0.12251333147287369,
|
| 57 |
+
"logits/rejected": 0.22839057445526123,
|
| 58 |
+
"logps/chosen": -353.02825927734375,
|
| 59 |
+
"logps/rejected": -329.4814147949219,
|
| 60 |
+
"loss": 0.3833,
|
| 61 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 62 |
+
"rewards/chosen": 0.00021826496231369674,
|
| 63 |
+
"rewards/margins": 0.0004608921299222857,
|
| 64 |
+
"rewards/rejected": -0.00024262710940092802,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1,
|
| 69 |
+
"learning_rate": 4.761904761904762e-06,
|
| 70 |
+
"logits/chosen": 0.10691192001104355,
|
| 71 |
+
"logits/rejected": 0.16547340154647827,
|
| 72 |
+
"logps/chosen": -330.2546691894531,
|
| 73 |
+
"logps/rejected": -321.59222412109375,
|
| 74 |
+
"loss": 0.3676,
|
| 75 |
+
"rewards/accuracies": 0.65625,
|
| 76 |
+
"rewards/chosen": 0.000455420755315572,
|
| 77 |
+
"rewards/margins": 0.0009521494503132999,
|
| 78 |
+
"rewards/rejected": -0.0004967286949977279,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12,
|
| 83 |
+
"learning_rate": 4.994357350311441e-06,
|
| 84 |
+
"logits/chosen": 0.11281980574131012,
|
| 85 |
+
"logits/rejected": 0.1559869647026062,
|
| 86 |
+
"logps/chosen": -333.9978942871094,
|
| 87 |
+
"logps/rejected": -305.3361511230469,
|
| 88 |
+
"loss": 0.3648,
|
| 89 |
+
"rewards/accuracies": 0.65625,
|
| 90 |
+
"rewards/chosen": 0.0016132129821926355,
|
| 91 |
+
"rewards/margins": 0.002727704355493188,
|
| 92 |
+
"rewards/rejected": -0.0011144911404699087,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.14,
|
| 97 |
+
"learning_rate": 4.97147773390341e-06,
|
| 98 |
+
"logits/chosen": 0.1588975489139557,
|
| 99 |
+
"logits/rejected": 0.22720813751220703,
|
| 100 |
+
"logps/chosen": -328.2370300292969,
|
| 101 |
+
"logps/rejected": -326.60650634765625,
|
| 102 |
+
"loss": 0.3615,
|
| 103 |
+
"rewards/accuracies": 0.699999988079071,
|
| 104 |
+
"rewards/chosen": 0.002356642857193947,
|
| 105 |
+
"rewards/margins": 0.006414002738893032,
|
| 106 |
+
"rewards/rejected": -0.004057359881699085,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.17,
|
| 111 |
+
"learning_rate": 4.931169703639282e-06,
|
| 112 |
+
"logits/chosen": 0.12001170963048935,
|
| 113 |
+
"logits/rejected": 0.2352440357208252,
|
| 114 |
+
"logps/chosen": -363.2486267089844,
|
| 115 |
+
"logps/rejected": -353.33795166015625,
|
| 116 |
+
"loss": 0.3571,
|
| 117 |
+
"rewards/accuracies": 0.71875,
|
| 118 |
+
"rewards/chosen": 0.001306799822486937,
|
| 119 |
+
"rewards/margins": 0.013176659122109413,
|
| 120 |
+
"rewards/rejected": -0.011869858019053936,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.19,
|
| 125 |
+
"learning_rate": 4.873717504456219e-06,
|
| 126 |
+
"logits/chosen": 0.12399880588054657,
|
| 127 |
+
"logits/rejected": 0.13322117924690247,
|
| 128 |
+
"logps/chosen": -351.8349304199219,
|
| 129 |
+
"logps/rejected": -346.0111999511719,
|
| 130 |
+
"loss": 0.3553,
|
| 131 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 132 |
+
"rewards/chosen": -0.01495267916470766,
|
| 133 |
+
"rewards/margins": 0.018836025148630142,
|
| 134 |
+
"rewards/rejected": -0.03378870338201523,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.22,
|
| 139 |
+
"learning_rate": 4.7995262788689865e-06,
|
| 140 |
+
"logits/chosen": 0.15232697129249573,
|
| 141 |
+
"logits/rejected": 0.16080796718597412,
|
| 142 |
+
"logps/chosen": -415.25408935546875,
|
| 143 |
+
"logps/rejected": -472.15838623046875,
|
| 144 |
+
"loss": 0.3288,
|
| 145 |
+
"rewards/accuracies": 0.699999988079071,
|
| 146 |
+
"rewards/chosen": -0.060072846710681915,
|
| 147 |
+
"rewards/margins": 0.05840995907783508,
|
| 148 |
+
"rewards/rejected": -0.118482805788517,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.24,
|
| 153 |
+
"learning_rate": 4.709119209978242e-06,
|
| 154 |
+
"logits/chosen": 0.13195478916168213,
|
| 155 |
+
"logits/rejected": 0.1627262532711029,
|
| 156 |
+
"logps/chosen": -528.2550659179688,
|
| 157 |
+
"logps/rejected": -582.0850219726562,
|
| 158 |
+
"loss": 0.3394,
|
| 159 |
+
"rewards/accuracies": 0.65625,
|
| 160 |
+
"rewards/chosen": -0.16017559170722961,
|
| 161 |
+
"rewards/margins": 0.08082611858844757,
|
| 162 |
+
"rewards/rejected": -0.241001695394516,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.26,
|
| 167 |
+
"learning_rate": 4.603133832077953e-06,
|
| 168 |
+
"logits/chosen": 0.07086379081010818,
|
| 169 |
+
"logits/rejected": 0.13402250409126282,
|
| 170 |
+
"logps/chosen": -478.28155517578125,
|
| 171 |
+
"logps/rejected": -536.5116577148438,
|
| 172 |
+
"loss": 0.3097,
|
| 173 |
+
"rewards/accuracies": 0.6875,
|
| 174 |
+
"rewards/chosen": -0.1429453045129776,
|
| 175 |
+
"rewards/margins": 0.08558131754398346,
|
| 176 |
+
"rewards/rejected": -0.22852663695812225,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.29,
|
| 181 |
+
"learning_rate": 4.482317534878901e-06,
|
| 182 |
+
"logits/chosen": 0.04285923391580582,
|
| 183 |
+
"logits/rejected": 0.1404789388179779,
|
| 184 |
+
"logps/chosen": -488.8091735839844,
|
| 185 |
+
"logps/rejected": -566.8741455078125,
|
| 186 |
+
"loss": 0.304,
|
| 187 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 188 |
+
"rewards/chosen": -0.165468230843544,
|
| 189 |
+
"rewards/margins": 0.09716635942459106,
|
| 190 |
+
"rewards/rejected": -0.2626345753669739,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.31,
|
| 195 |
+
"learning_rate": 4.3475222930516484e-06,
|
| 196 |
+
"logits/chosen": 0.0852217823266983,
|
| 197 |
+
"logits/rejected": 0.12218357622623444,
|
| 198 |
+
"logps/chosen": -495.8277282714844,
|
| 199 |
+
"logps/rejected": -630.5374145507812,
|
| 200 |
+
"loss": 0.305,
|
| 201 |
+
"rewards/accuracies": 0.675000011920929,
|
| 202 |
+
"rewards/chosen": -0.18193328380584717,
|
| 203 |
+
"rewards/margins": 0.1324920505285263,
|
| 204 |
+
"rewards/rejected": -0.3144252896308899,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.34,
|
| 209 |
+
"learning_rate": 4.199698658255298e-06,
|
| 210 |
+
"logits/chosen": 0.10588987171649933,
|
| 211 |
+
"logits/rejected": 0.13151074945926666,
|
| 212 |
+
"logps/chosen": -534.0150146484375,
|
| 213 |
+
"logps/rejected": -573.1961059570312,
|
| 214 |
+
"loss": 0.3174,
|
| 215 |
+
"rewards/accuracies": 0.612500011920929,
|
| 216 |
+
"rewards/chosen": -0.1705319881439209,
|
| 217 |
+
"rewards/margins": 0.08429961651563644,
|
| 218 |
+
"rewards/rejected": -0.25483161211013794,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.36,
|
| 223 |
+
"learning_rate": 4.039889056019159e-06,
|
| 224 |
+
"logits/chosen": 0.14604777097702026,
|
| 225 |
+
"logits/rejected": 0.17551526427268982,
|
| 226 |
+
"logps/chosen": -508.1181640625,
|
| 227 |
+
"logps/rejected": -601.8496704101562,
|
| 228 |
+
"loss": 0.2995,
|
| 229 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 230 |
+
"rewards/chosen": -0.15790864825248718,
|
| 231 |
+
"rewards/margins": 0.10467412322759628,
|
| 232 |
+
"rewards/rejected": -0.2625827491283417,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.38,
|
| 237 |
+
"learning_rate": 3.869220434746509e-06,
|
| 238 |
+
"logits/chosen": 0.11204711347818375,
|
| 239 |
+
"logits/rejected": 0.12582647800445557,
|
| 240 |
+
"logps/chosen": -511.799072265625,
|
| 241 |
+
"logps/rejected": -630.4769287109375,
|
| 242 |
+
"loss": 0.3083,
|
| 243 |
+
"rewards/accuracies": 0.71875,
|
| 244 |
+
"rewards/chosen": -0.17171703279018402,
|
| 245 |
+
"rewards/margins": 0.12071826308965683,
|
| 246 |
+
"rewards/rejected": -0.29243525862693787,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.41,
|
| 251 |
+
"learning_rate": 3.688896318678322e-06,
|
| 252 |
+
"logits/chosen": 0.08149586617946625,
|
| 253 |
+
"logits/rejected": 0.1699526309967041,
|
| 254 |
+
"logps/chosen": -502.21331787109375,
|
| 255 |
+
"logps/rejected": -597.2528076171875,
|
| 256 |
+
"loss": 0.3003,
|
| 257 |
+
"rewards/accuracies": 0.78125,
|
| 258 |
+
"rewards/chosen": -0.1475558578968048,
|
| 259 |
+
"rewards/margins": 0.137506365776062,
|
| 260 |
+
"rewards/rejected": -0.28506219387054443,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.43,
|
| 265 |
+
"learning_rate": 3.5001883208580668e-06,
|
| 266 |
+
"logits/chosen": 0.06874342262744904,
|
| 267 |
+
"logits/rejected": 0.17765206098556519,
|
| 268 |
+
"logps/chosen": -517.6918334960938,
|
| 269 |
+
"logps/rejected": -670.2382202148438,
|
| 270 |
+
"loss": 0.2989,
|
| 271 |
+
"rewards/accuracies": 0.668749988079071,
|
| 272 |
+
"rewards/chosen": -0.18062789738178253,
|
| 273 |
+
"rewards/margins": 0.14384225010871887,
|
| 274 |
+
"rewards/rejected": -0.3244701325893402,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.46,
|
| 279 |
+
"learning_rate": 3.30442717594657e-06,
|
| 280 |
+
"logits/chosen": 0.08455907553434372,
|
| 281 |
+
"logits/rejected": 0.16155509650707245,
|
| 282 |
+
"logps/chosen": -573.642333984375,
|
| 283 |
+
"logps/rejected": -693.3612060546875,
|
| 284 |
+
"loss": 0.2908,
|
| 285 |
+
"rewards/accuracies": 0.637499988079071,
|
| 286 |
+
"rewards/chosen": -0.21603624522686005,
|
| 287 |
+
"rewards/margins": 0.11141183227300644,
|
| 288 |
+
"rewards/rejected": -0.3274480700492859,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.48,
|
| 293 |
+
"learning_rate": 3.102993356121938e-06,
|
| 294 |
+
"logits/chosen": 0.13803192973136902,
|
| 295 |
+
"logits/rejected": 0.13260796666145325,
|
| 296 |
+
"logps/chosen": -504.03070068359375,
|
| 297 |
+
"logps/rejected": -662.9249267578125,
|
| 298 |
+
"loss": 0.2837,
|
| 299 |
+
"rewards/accuracies": 0.65625,
|
| 300 |
+
"rewards/chosen": -0.20684099197387695,
|
| 301 |
+
"rewards/margins": 0.132938414812088,
|
| 302 |
+
"rewards/rejected": -0.33977940678596497,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5,
|
| 307 |
+
"learning_rate": 2.8973073362395e-06,
|
| 308 |
+
"logits/chosen": 0.1248478889465332,
|
| 309 |
+
"logits/rejected": 0.18150393664836884,
|
| 310 |
+
"logps/chosen": -556.743896484375,
|
| 311 |
+
"logps/rejected": -682.061279296875,
|
| 312 |
+
"loss": 0.3043,
|
| 313 |
+
"rewards/accuracies": 0.699999988079071,
|
| 314 |
+
"rewards/chosen": -0.1968514621257782,
|
| 315 |
+
"rewards/margins": 0.12893599271774292,
|
| 316 |
+
"rewards/rejected": -0.3257874846458435,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.53,
|
| 321 |
+
"learning_rate": 2.6888195769001147e-06,
|
| 322 |
+
"logits/chosen": 0.12090057134628296,
|
| 323 |
+
"logits/rejected": 0.16103163361549377,
|
| 324 |
+
"logps/chosen": -576.2421875,
|
| 325 |
+
"logps/rejected": -636.4088134765625,
|
| 326 |
+
"loss": 0.2942,
|
| 327 |
+
"rewards/accuracies": 0.706250011920929,
|
| 328 |
+
"rewards/chosen": -0.17859044671058655,
|
| 329 |
+
"rewards/margins": 0.11207801103591919,
|
| 330 |
+
"rewards/rejected": -0.29066842794418335,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.55,
|
| 335 |
+
"learning_rate": 2.479000296064417e-06,
|
| 336 |
+
"logits/chosen": 0.11349457502365112,
|
| 337 |
+
"logits/rejected": 0.17926748096942902,
|
| 338 |
+
"logps/chosen": -535.904052734375,
|
| 339 |
+
"logps/rejected": -646.5347900390625,
|
| 340 |
+
"loss": 0.3018,
|
| 341 |
+
"rewards/accuracies": 0.699999988079071,
|
| 342 |
+
"rewards/chosen": -0.18949127197265625,
|
| 343 |
+
"rewards/margins": 0.14192432165145874,
|
| 344 |
+
"rewards/rejected": -0.3314156234264374,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.58,
|
| 349 |
+
"learning_rate": 2.269329101341745e-06,
|
| 350 |
+
"logits/chosen": 0.08516987413167953,
|
| 351 |
+
"logits/rejected": 0.14970402419567108,
|
| 352 |
+
"logps/chosen": -517.2757568359375,
|
| 353 |
+
"logps/rejected": -607.4237060546875,
|
| 354 |
+
"loss": 0.2892,
|
| 355 |
+
"rewards/accuracies": 0.643750011920929,
|
| 356 |
+
"rewards/chosen": -0.17799392342567444,
|
| 357 |
+
"rewards/margins": 0.11533834785223007,
|
| 358 |
+
"rewards/rejected": -0.2933322489261627,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6,
|
| 363 |
+
"learning_rate": 2.06128455606496e-06,
|
| 364 |
+
"logits/chosen": 0.08016245067119598,
|
| 365 |
+
"logits/rejected": 0.16319520771503448,
|
| 366 |
+
"logps/chosen": -497.9122619628906,
|
| 367 |
+
"logps/rejected": -589.7242431640625,
|
| 368 |
+
"loss": 0.3044,
|
| 369 |
+
"rewards/accuracies": 0.65625,
|
| 370 |
+
"rewards/chosen": -0.19338397681713104,
|
| 371 |
+
"rewards/margins": 0.09803850203752518,
|
| 372 |
+
"rewards/rejected": -0.2914224863052368,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.62,
|
| 377 |
+
"learning_rate": 1.856333752729311e-06,
|
| 378 |
+
"logits/chosen": 0.15101440250873566,
|
| 379 |
+
"logits/rejected": 0.171871617436409,
|
| 380 |
+
"logps/chosen": -571.2918090820312,
|
| 381 |
+
"logps/rejected": -696.5802001953125,
|
| 382 |
+
"loss": 0.3184,
|
| 383 |
+
"rewards/accuracies": 0.71875,
|
| 384 |
+
"rewards/chosen": -0.20878608524799347,
|
| 385 |
+
"rewards/margins": 0.13470368087291718,
|
| 386 |
+
"rewards/rejected": -0.34348976612091064,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.65,
|
| 391 |
+
"learning_rate": 1.6559219673215784e-06,
|
| 392 |
+
"logits/chosen": 0.13930802047252655,
|
| 393 |
+
"logits/rejected": 0.20976956188678741,
|
| 394 |
+
"logps/chosen": -531.32177734375,
|
| 395 |
+
"logps/rejected": -613.5001220703125,
|
| 396 |
+
"loss": 0.2899,
|
| 397 |
+
"rewards/accuracies": 0.668749988079071,
|
| 398 |
+
"rewards/chosen": -0.17231304943561554,
|
| 399 |
+
"rewards/margins": 0.12125241756439209,
|
| 400 |
+
"rewards/rejected": -0.29356545209884644,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.67,
|
| 405 |
+
"learning_rate": 1.4614624674952843e-06,
|
| 406 |
+
"logits/chosen": 0.11396761983633041,
|
| 407 |
+
"logits/rejected": 0.2314365655183792,
|
| 408 |
+
"logps/chosen": -552.0916748046875,
|
| 409 |
+
"logps/rejected": -652.0897827148438,
|
| 410 |
+
"loss": 0.2926,
|
| 411 |
+
"rewards/accuracies": 0.706250011920929,
|
| 412 |
+
"rewards/chosen": -0.18660280108451843,
|
| 413 |
+
"rewards/margins": 0.11572613567113876,
|
| 414 |
+
"rewards/rejected": -0.3023289442062378,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7,
|
| 419 |
+
"learning_rate": 1.2743265464628787e-06,
|
| 420 |
+
"logits/chosen": 0.06689377129077911,
|
| 421 |
+
"logits/rejected": 0.1653607338666916,
|
| 422 |
+
"logps/chosen": -559.4171142578125,
|
| 423 |
+
"logps/rejected": -663.8862915039062,
|
| 424 |
+
"loss": 0.2877,
|
| 425 |
+
"rewards/accuracies": 0.699999988079071,
|
| 426 |
+
"rewards/chosen": -0.19495370984077454,
|
| 427 |
+
"rewards/margins": 0.13509666919708252,
|
| 428 |
+
"rewards/rejected": -0.33005034923553467,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.72,
|
| 433 |
+
"learning_rate": 1.0958338528840893e-06,
|
| 434 |
+
"logits/chosen": 0.0707981139421463,
|
| 435 |
+
"logits/rejected": 0.19608157873153687,
|
| 436 |
+
"logps/chosen": -585.0418090820312,
|
| 437 |
+
"logps/rejected": -708.06689453125,
|
| 438 |
+
"loss": 0.2956,
|
| 439 |
+
"rewards/accuracies": 0.71875,
|
| 440 |
+
"rewards/chosen": -0.21173438429832458,
|
| 441 |
+
"rewards/margins": 0.14347293972969055,
|
| 442 |
+
"rewards/rejected": -0.3552073538303375,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.74,
|
| 447 |
+
"learning_rate": 9.272430849423175e-07,
|
| 448 |
+
"logits/chosen": 0.12051234394311905,
|
| 449 |
+
"logits/rejected": 0.22985529899597168,
|
| 450 |
+
"logps/chosen": -581.6112670898438,
|
| 451 |
+
"logps/rejected": -700.2015380859375,
|
| 452 |
+
"loss": 0.2826,
|
| 453 |
+
"rewards/accuracies": 0.731249988079071,
|
| 454 |
+
"rewards/chosen": -0.1966112107038498,
|
| 455 |
+
"rewards/margins": 0.1535426378250122,
|
| 456 |
+
"rewards/rejected": -0.3501538634300232,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.77,
|
| 461 |
+
"learning_rate": 7.697431142327633e-07,
|
| 462 |
+
"logits/chosen": 0.14221158623695374,
|
| 463 |
+
"logits/rejected": 0.16759946942329407,
|
| 464 |
+
"logps/chosen": -543.4645385742188,
|
| 465 |
+
"logps/rejected": -653.2728271484375,
|
| 466 |
+
"loss": 0.2983,
|
| 467 |
+
"rewards/accuracies": 0.706250011920929,
|
| 468 |
+
"rewards/chosen": -0.19766229391098022,
|
| 469 |
+
"rewards/margins": 0.13313183188438416,
|
| 470 |
+
"rewards/rejected": -0.33079415559768677,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.79,
|
| 475 |
+
"learning_rate": 6.244446020550182e-07,
|
| 476 |
+
"logits/chosen": 0.1523372381925583,
|
| 477 |
+
"logits/rejected": 0.17464013397693634,
|
| 478 |
+
"logps/chosen": -559.0565185546875,
|
| 479 |
+
"logps/rejected": -647.8815307617188,
|
| 480 |
+
"loss": 0.2835,
|
| 481 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 482 |
+
"rewards/chosen": -0.20685505867004395,
|
| 483 |
+
"rewards/margins": 0.11294861137866974,
|
| 484 |
+
"rewards/rejected": -0.3198036551475525,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.82,
|
| 489 |
+
"learning_rate": 4.923721672305148e-07,
|
| 490 |
+
"logits/chosen": 0.07727678120136261,
|
| 491 |
+
"logits/rejected": 0.14975441992282867,
|
| 492 |
+
"logps/chosen": -530.6942138671875,
|
| 493 |
+
"logps/rejected": -737.3821411132812,
|
| 494 |
+
"loss": 0.2821,
|
| 495 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 496 |
+
"rewards/chosen": -0.1940075010061264,
|
| 497 |
+
"rewards/margins": 0.18392714858055115,
|
| 498 |
+
"rewards/rejected": -0.37793463468551636,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.84,
|
| 503 |
+
"learning_rate": 3.7445716067596506e-07,
|
| 504 |
+
"logits/chosen": 0.06651227176189423,
|
| 505 |
+
"logits/rejected": 0.1357874870300293,
|
| 506 |
+
"logps/chosen": -568.581298828125,
|
| 507 |
+
"logps/rejected": -677.1904907226562,
|
| 508 |
+
"loss": 0.2696,
|
| 509 |
+
"rewards/accuracies": 0.699999988079071,
|
| 510 |
+
"rewards/chosen": -0.22338159382343292,
|
| 511 |
+
"rewards/margins": 0.14005926251411438,
|
| 512 |
+
"rewards/rejected": -0.3634408414363861,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.86,
|
| 517 |
+
"learning_rate": 2.7153109768518926e-07,
|
| 518 |
+
"logits/chosen": 0.12754273414611816,
|
| 519 |
+
"logits/rejected": 0.18894067406654358,
|
| 520 |
+
"logps/chosen": -551.6950073242188,
|
| 521 |
+
"logps/rejected": -662.9625244140625,
|
| 522 |
+
"loss": 0.2798,
|
| 523 |
+
"rewards/accuracies": 0.6875,
|
| 524 |
+
"rewards/chosen": -0.20485155284404755,
|
| 525 |
+
"rewards/margins": 0.13481572270393372,
|
| 526 |
+
"rewards/rejected": -0.33966726064682007,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.89,
|
| 531 |
+
"learning_rate": 1.8431979423369607e-07,
|
| 532 |
+
"logits/chosen": 0.1290164738893509,
|
| 533 |
+
"logits/rejected": 0.2077597826719284,
|
| 534 |
+
"logps/chosen": -539.3775634765625,
|
| 535 |
+
"logps/rejected": -648.92138671875,
|
| 536 |
+
"loss": 0.3018,
|
| 537 |
+
"rewards/accuracies": 0.625,
|
| 538 |
+
"rewards/chosen": -0.22309288382530212,
|
| 539 |
+
"rewards/margins": 0.11153991520404816,
|
| 540 |
+
"rewards/rejected": -0.3346328139305115,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.91,
|
| 545 |
+
"learning_rate": 1.1343824865573422e-07,
|
| 546 |
+
"logits/chosen": 0.12593218684196472,
|
| 547 |
+
"logits/rejected": 0.1743309050798416,
|
| 548 |
+
"logps/chosen": -534.8792114257812,
|
| 549 |
+
"logps/rejected": -619.5510864257812,
|
| 550 |
+
"loss": 0.3046,
|
| 551 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 552 |
+
"rewards/chosen": -0.2254379242658615,
|
| 553 |
+
"rewards/margins": 0.10715726763010025,
|
| 554 |
+
"rewards/rejected": -0.33259516954421997,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.94,
|
| 559 |
+
"learning_rate": 5.9386304787299175e-08,
|
| 560 |
+
"logits/chosen": 0.15326878428459167,
|
| 561 |
+
"logits/rejected": 0.19667108356952667,
|
| 562 |
+
"logps/chosen": -596.8463745117188,
|
| 563 |
+
"logps/rejected": -687.6807861328125,
|
| 564 |
+
"loss": 0.2838,
|
| 565 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 566 |
+
"rewards/chosen": -0.2249801903963089,
|
| 567 |
+
"rewards/margins": 0.12209514528512955,
|
| 568 |
+
"rewards/rejected": -0.34707534313201904,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.96,
|
| 573 |
+
"learning_rate": 2.2545127157831416e-08,
|
| 574 |
+
"logits/chosen": 0.08709342032670975,
|
| 575 |
+
"logits/rejected": 0.12705549597740173,
|
| 576 |
+
"logps/chosen": -551.4010620117188,
|
| 577 |
+
"logps/rejected": -672.7706298828125,
|
| 578 |
+
"loss": 0.2969,
|
| 579 |
+
"rewards/accuracies": 0.675000011920929,
|
| 580 |
+
"rewards/chosen": -0.2128075808286667,
|
| 581 |
+
"rewards/margins": 0.12216651439666748,
|
| 582 |
+
"rewards/rejected": -0.33497413992881775,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.98,
|
| 587 |
+
"learning_rate": 3.1745130869123564e-09,
|
| 588 |
+
"logits/chosen": 0.07737437635660172,
|
| 589 |
+
"logits/rejected": 0.13851606845855713,
|
| 590 |
+
"logps/chosen": -584.75,
|
| 591 |
+
"logps/rejected": -722.1080322265625,
|
| 592 |
+
"loss": 0.2947,
|
| 593 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 594 |
+
"rewards/chosen": -0.22851315140724182,
|
| 595 |
+
"rewards/margins": 0.13729806244373322,
|
| 596 |
+
"rewards/rejected": -0.36581122875213623,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.0,
|
| 601 |
+
"step": 416,
|
| 602 |
+
"total_flos": 0.0,
|
| 603 |
+
"train_loss": 0.3107354554276054,
|
| 604 |
+
"train_runtime": 5919.6286,
|
| 605 |
+
"train_samples_per_second": 3.379,
|
| 606 |
+
"train_steps_per_second": 0.07
|
| 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 |
+
}
|