Model save
Browse files- README.md +62 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +387 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: meta-llama/Meta-Llama-3-8B
|
| 3 |
+
library_name: peft
|
| 4 |
+
license: llama3
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- dpo
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
model-index:
|
| 10 |
+
- name: llama3-sudo-dpo-no-instruct
|
| 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 |
+
# llama3-sudo-dpo-no-instruct
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on an unknown 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: 2
|
| 40 |
+
- eval_batch_size: 8
|
| 41 |
+
- seed: 42
|
| 42 |
+
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 8
|
| 44 |
+
- gradient_accumulation_steps: 4
|
| 45 |
+
- total_train_batch_size: 64
|
| 46 |
+
- total_eval_batch_size: 64
|
| 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 0.12.0
|
| 59 |
+
- Transformers 4.44.0
|
| 60 |
+
- Pytorch 2.1.2
|
| 61 |
+
- Datasets 2.20.0
|
| 62 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.15411589154528807,
|
| 5 |
+
"train_runtime": 2209.5741,
|
| 6 |
+
"train_samples": 14642,
|
| 7 |
+
"train_samples_per_second": 6.627,
|
| 8 |
+
"train_steps_per_second": 0.104
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.15411589154528807,
|
| 5 |
+
"train_runtime": 2209.5741,
|
| 6 |
+
"train_samples": 14642,
|
| 7 |
+
"train_samples_per_second": 6.627,
|
| 8 |
+
"train_steps_per_second": 0.104
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,387 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 229,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004366812227074236,
|
| 13 |
+
"grad_norm": 0.3204378540751742,
|
| 14 |
+
"learning_rate": 2.173913043478261e-07,
|
| 15 |
+
"logits/chosen": -1.3045543432235718,
|
| 16 |
+
"logits/rejected": -1.2643918991088867,
|
| 17 |
+
"logps/chosen": -239.7493896484375,
|
| 18 |
+
"logps/rejected": -242.68679809570312,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.043668122270742356,
|
| 28 |
+
"grad_norm": 0.3460819923447899,
|
| 29 |
+
"learning_rate": 2.173913043478261e-06,
|
| 30 |
+
"logits/chosen": -1.25380539894104,
|
| 31 |
+
"logits/rejected": -1.260703682899475,
|
| 32 |
+
"logps/chosen": -256.6331481933594,
|
| 33 |
+
"logps/rejected": -223.29080200195312,
|
| 34 |
+
"loss": 0.6929,
|
| 35 |
+
"rewards/accuracies": 0.4027777910232544,
|
| 36 |
+
"rewards/chosen": 0.0002949640329461545,
|
| 37 |
+
"rewards/margins": -0.0002991040237247944,
|
| 38 |
+
"rewards/rejected": 0.0005940680857747793,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.08733624454148471,
|
| 43 |
+
"grad_norm": 0.39683791430414567,
|
| 44 |
+
"learning_rate": 4.347826086956522e-06,
|
| 45 |
+
"logits/chosen": -1.2746273279190063,
|
| 46 |
+
"logits/rejected": -1.2591041326522827,
|
| 47 |
+
"logps/chosen": -243.3075714111328,
|
| 48 |
+
"logps/rejected": -275.537109375,
|
| 49 |
+
"loss": 0.6902,
|
| 50 |
+
"rewards/accuracies": 0.699999988079071,
|
| 51 |
+
"rewards/chosen": 0.006293324287980795,
|
| 52 |
+
"rewards/margins": 0.006268353201448917,
|
| 53 |
+
"rewards/rejected": 2.497106288501527e-05,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.13100436681222707,
|
| 58 |
+
"grad_norm": 0.5419622183450867,
|
| 59 |
+
"learning_rate": 4.985768230048011e-06,
|
| 60 |
+
"logits/chosen": -1.2166337966918945,
|
| 61 |
+
"logits/rejected": -1.221422791481018,
|
| 62 |
+
"logps/chosen": -242.04238891601562,
|
| 63 |
+
"logps/rejected": -227.62631225585938,
|
| 64 |
+
"loss": 0.6769,
|
| 65 |
+
"rewards/accuracies": 0.887499988079071,
|
| 66 |
+
"rewards/chosen": 0.014850045554339886,
|
| 67 |
+
"rewards/margins": 0.032028842717409134,
|
| 68 |
+
"rewards/rejected": -0.017178794369101524,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.17467248908296942,
|
| 73 |
+
"grad_norm": 0.7063163692451463,
|
| 74 |
+
"learning_rate": 4.9164513914144005e-06,
|
| 75 |
+
"logits/chosen": -1.188439965248108,
|
| 76 |
+
"logits/rejected": -1.2070634365081787,
|
| 77 |
+
"logps/chosen": -238.8345184326172,
|
| 78 |
+
"logps/rejected": -241.16683959960938,
|
| 79 |
+
"loss": 0.6383,
|
| 80 |
+
"rewards/accuracies": 1.0,
|
| 81 |
+
"rewards/chosen": 0.02510463260114193,
|
| 82 |
+
"rewards/margins": 0.11083698272705078,
|
| 83 |
+
"rewards/rejected": -0.0857323482632637,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.2183406113537118,
|
| 88 |
+
"grad_norm": 1.1679104233394753,
|
| 89 |
+
"learning_rate": 4.791042480696179e-06,
|
| 90 |
+
"logits/chosen": -1.1544982194900513,
|
| 91 |
+
"logits/rejected": -1.1601812839508057,
|
| 92 |
+
"logps/chosen": -220.57308959960938,
|
| 93 |
+
"logps/rejected": -297.6596984863281,
|
| 94 |
+
"loss": 0.522,
|
| 95 |
+
"rewards/accuracies": 1.0,
|
| 96 |
+
"rewards/chosen": 0.07746505737304688,
|
| 97 |
+
"rewards/margins": 0.39530041813850403,
|
| 98 |
+
"rewards/rejected": -0.31783536076545715,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.26200873362445415,
|
| 103 |
+
"grad_norm": 1.6679501998965438,
|
| 104 |
+
"learning_rate": 4.612452562309975e-06,
|
| 105 |
+
"logits/chosen": -0.9797567129135132,
|
| 106 |
+
"logits/rejected": -0.8914886713027954,
|
| 107 |
+
"logps/chosen": -260.34539794921875,
|
| 108 |
+
"logps/rejected": -361.8608093261719,
|
| 109 |
+
"loss": 0.2727,
|
| 110 |
+
"rewards/accuracies": 1.0,
|
| 111 |
+
"rewards/chosen": -0.15145885944366455,
|
| 112 |
+
"rewards/margins": 1.2886455059051514,
|
| 113 |
+
"rewards/rejected": -1.440104365348816,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3056768558951965,
|
| 118 |
+
"grad_norm": 2.030171468208308,
|
| 119 |
+
"learning_rate": 4.384827169085993e-06,
|
| 120 |
+
"logits/chosen": -0.41109657287597656,
|
| 121 |
+
"logits/rejected": 0.050916582345962524,
|
| 122 |
+
"logps/chosen": -387.8983459472656,
|
| 123 |
+
"logps/rejected": -1060.7069091796875,
|
| 124 |
+
"loss": 0.0287,
|
| 125 |
+
"rewards/accuracies": 1.0,
|
| 126 |
+
"rewards/chosen": -1.4819408655166626,
|
| 127 |
+
"rewards/margins": 6.849398612976074,
|
| 128 |
+
"rewards/rejected": -8.331339836120605,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.34934497816593885,
|
| 133 |
+
"grad_norm": 0.5567289395039918,
|
| 134 |
+
"learning_rate": 4.1134500737541026e-06,
|
| 135 |
+
"logits/chosen": 0.08877621591091156,
|
| 136 |
+
"logits/rejected": 1.3163907527923584,
|
| 137 |
+
"logps/chosen": -386.9926452636719,
|
| 138 |
+
"logps/rejected": -2022.4027099609375,
|
| 139 |
+
"loss": 0.0013,
|
| 140 |
+
"rewards/accuracies": 1.0,
|
| 141 |
+
"rewards/chosen": -1.6194961071014404,
|
| 142 |
+
"rewards/margins": 16.02116584777832,
|
| 143 |
+
"rewards/rejected": -17.64065933227539,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.3930131004366812,
|
| 148 |
+
"grad_norm": 0.16624156411199678,
|
| 149 |
+
"learning_rate": 3.8046206389447916e-06,
|
| 150 |
+
"logits/chosen": -0.11295245587825775,
|
| 151 |
+
"logits/rejected": 1.0744993686676025,
|
| 152 |
+
"logps/chosen": -369.5999450683594,
|
| 153 |
+
"logps/rejected": -1938.8245849609375,
|
| 154 |
+
"loss": 0.002,
|
| 155 |
+
"rewards/accuracies": 1.0,
|
| 156 |
+
"rewards/chosen": -1.2554817199707031,
|
| 157 |
+
"rewards/margins": 15.89024543762207,
|
| 158 |
+
"rewards/rejected": -17.145727157592773,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.4366812227074236,
|
| 163 |
+
"grad_norm": 0.0037156101519251455,
|
| 164 |
+
"learning_rate": 3.4655075927279576e-06,
|
| 165 |
+
"logits/chosen": 0.12383558601140976,
|
| 166 |
+
"logits/rejected": 1.5539542436599731,
|
| 167 |
+
"logps/chosen": -325.15045166015625,
|
| 168 |
+
"logps/rejected": -2191.79248046875,
|
| 169 |
+
"loss": 0.0,
|
| 170 |
+
"rewards/accuracies": 1.0,
|
| 171 |
+
"rewards/chosen": -1.0599058866500854,
|
| 172 |
+
"rewards/margins": 18.508365631103516,
|
| 173 |
+
"rewards/rejected": -19.56827163696289,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.48034934497816595,
|
| 178 |
+
"grad_norm": 0.003926929804724353,
|
| 179 |
+
"learning_rate": 3.1039826239365754e-06,
|
| 180 |
+
"logits/chosen": 0.1838320940732956,
|
| 181 |
+
"logits/rejected": 1.4217793941497803,
|
| 182 |
+
"logps/chosen": -386.07891845703125,
|
| 183 |
+
"logps/rejected": -2162.380615234375,
|
| 184 |
+
"loss": 0.0017,
|
| 185 |
+
"rewards/accuracies": 1.0,
|
| 186 |
+
"rewards/chosen": -1.3692742586135864,
|
| 187 |
+
"rewards/margins": 17.987213134765625,
|
| 188 |
+
"rewards/rejected": -19.35648536682129,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.5240174672489083,
|
| 193 |
+
"grad_norm": 0.006662495713706298,
|
| 194 |
+
"learning_rate": 2.7284376599571776e-06,
|
| 195 |
+
"logits/chosen": -0.09864537417888641,
|
| 196 |
+
"logits/rejected": 1.3508899211883545,
|
| 197 |
+
"logps/chosen": -324.0560302734375,
|
| 198 |
+
"logps/rejected": -2175.47412109375,
|
| 199 |
+
"loss": 0.0003,
|
| 200 |
+
"rewards/accuracies": 1.0,
|
| 201 |
+
"rewards/chosen": -0.735422670841217,
|
| 202 |
+
"rewards/margins": 18.709741592407227,
|
| 203 |
+
"rewards/rejected": -19.44516372680664,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.5676855895196506,
|
| 208 |
+
"grad_norm": 0.000691033289700338,
|
| 209 |
+
"learning_rate": 2.3475900684411027e-06,
|
| 210 |
+
"logits/chosen": -0.16790446639060974,
|
| 211 |
+
"logits/rejected": 1.2784119844436646,
|
| 212 |
+
"logps/chosen": -324.835205078125,
|
| 213 |
+
"logps/rejected": -2195.280029296875,
|
| 214 |
+
"loss": 0.0001,
|
| 215 |
+
"rewards/accuracies": 1.0,
|
| 216 |
+
"rewards/chosen": -0.8028841018676758,
|
| 217 |
+
"rewards/margins": 18.801555633544922,
|
| 218 |
+
"rewards/rejected": -19.604440689086914,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.611353711790393,
|
| 223 |
+
"grad_norm": 0.0005479383695858241,
|
| 224 |
+
"learning_rate": 1.970280304707447e-06,
|
| 225 |
+
"logits/chosen": -0.232609823346138,
|
| 226 |
+
"logits/rejected": 1.35557222366333,
|
| 227 |
+
"logps/chosen": -294.3355407714844,
|
| 228 |
+
"logps/rejected": -2262.7880859375,
|
| 229 |
+
"loss": 0.0012,
|
| 230 |
+
"rewards/accuracies": 1.0,
|
| 231 |
+
"rewards/chosen": -0.6753333210945129,
|
| 232 |
+
"rewards/margins": 19.330249786376953,
|
| 233 |
+
"rewards/rejected": -20.005586624145508,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6550218340611353,
|
| 238 |
+
"grad_norm": 0.024069992589533133,
|
| 239 |
+
"learning_rate": 1.6052667019636462e-06,
|
| 240 |
+
"logits/chosen": -0.20313453674316406,
|
| 241 |
+
"logits/rejected": 1.1127307415008545,
|
| 242 |
+
"logps/chosen": -327.80712890625,
|
| 243 |
+
"logps/rejected": -2066.748779296875,
|
| 244 |
+
"loss": 0.0003,
|
| 245 |
+
"rewards/accuracies": 1.0,
|
| 246 |
+
"rewards/chosen": -0.8296041488647461,
|
| 247 |
+
"rewards/margins": 17.718467712402344,
|
| 248 |
+
"rewards/rejected": -18.548070907592773,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.6986899563318777,
|
| 253 |
+
"grad_norm": 0.0013866889384772708,
|
| 254 |
+
"learning_rate": 1.261022167792161e-06,
|
| 255 |
+
"logits/chosen": -0.3580602705478668,
|
| 256 |
+
"logits/rejected": 1.0687057971954346,
|
| 257 |
+
"logps/chosen": -308.1361083984375,
|
| 258 |
+
"logps/rejected": -2346.19091796875,
|
| 259 |
+
"loss": 0.0001,
|
| 260 |
+
"rewards/accuracies": 1.0,
|
| 261 |
+
"rewards/chosen": -0.6503447890281677,
|
| 262 |
+
"rewards/margins": 20.15079116821289,
|
| 263 |
+
"rewards/rejected": -20.801136016845703,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.74235807860262,
|
| 268 |
+
"grad_norm": 0.04572961498583179,
|
| 269 |
+
"learning_rate": 9.455375061024319e-07,
|
| 270 |
+
"logits/chosen": -0.18205393850803375,
|
| 271 |
+
"logits/rejected": 1.1636852025985718,
|
| 272 |
+
"logps/chosen": -326.67071533203125,
|
| 273 |
+
"logps/rejected": -2068.658935546875,
|
| 274 |
+
"loss": 0.0002,
|
| 275 |
+
"rewards/accuracies": 1.0,
|
| 276 |
+
"rewards/chosen": -0.7653212547302246,
|
| 277 |
+
"rewards/margins": 17.772567749023438,
|
| 278 |
+
"rewards/rejected": -18.537891387939453,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.7860262008733624,
|
| 283 |
+
"grad_norm": 0.016866022956513082,
|
| 284 |
+
"learning_rate": 6.661359299530626e-07,
|
| 285 |
+
"logits/chosen": -0.15331922471523285,
|
| 286 |
+
"logits/rejected": 1.3502165079116821,
|
| 287 |
+
"logps/chosen": -307.6412658691406,
|
| 288 |
+
"logps/rejected": -2298.45751953125,
|
| 289 |
+
"loss": 0.0001,
|
| 290 |
+
"rewards/accuracies": 1.0,
|
| 291 |
+
"rewards/chosen": -0.7325421571731567,
|
| 292 |
+
"rewards/margins": 19.838850021362305,
|
| 293 |
+
"rewards/rejected": -20.571393966674805,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.8296943231441049,
|
| 298 |
+
"grad_norm": 0.0007817495246332639,
|
| 299 |
+
"learning_rate": 4.293030708802834e-07,
|
| 300 |
+
"logits/chosen": -0.1974346786737442,
|
| 301 |
+
"logits/rejected": 1.3936047554016113,
|
| 302 |
+
"logps/chosen": -320.288330078125,
|
| 303 |
+
"logps/rejected": -2285.126220703125,
|
| 304 |
+
"loss": 0.0001,
|
| 305 |
+
"rewards/accuracies": 1.0,
|
| 306 |
+
"rewards/chosen": -0.7533941864967346,
|
| 307 |
+
"rewards/margins": 19.595779418945312,
|
| 308 |
+
"rewards/rejected": -20.349172592163086,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.8733624454148472,
|
| 313 |
+
"grad_norm": 9.852312248548991e-05,
|
| 314 |
+
"learning_rate": 2.405364306547955e-07,
|
| 315 |
+
"logits/chosen": -0.02749384008347988,
|
| 316 |
+
"logits/rejected": 1.651100516319275,
|
| 317 |
+
"logps/chosen": -301.0024108886719,
|
| 318 |
+
"logps/rejected": -2246.86474609375,
|
| 319 |
+
"loss": 0.0001,
|
| 320 |
+
"rewards/accuracies": 1.0,
|
| 321 |
+
"rewards/chosen": -0.6887163519859314,
|
| 322 |
+
"rewards/margins": 19.42214012145996,
|
| 323 |
+
"rewards/rejected": -20.110858917236328,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.9170305676855895,
|
| 328 |
+
"grad_norm": 0.00601105052658747,
|
| 329 |
+
"learning_rate": 1.0421777008019663e-07,
|
| 330 |
+
"logits/chosen": -0.15921764075756073,
|
| 331 |
+
"logits/rejected": 1.3648946285247803,
|
| 332 |
+
"logps/chosen": -329.92120361328125,
|
| 333 |
+
"logps/rejected": -2383.26416015625,
|
| 334 |
+
"loss": 0.0001,
|
| 335 |
+
"rewards/accuracies": 1.0,
|
| 336 |
+
"rewards/chosen": -0.8647125363349915,
|
| 337 |
+
"rewards/margins": 20.413026809692383,
|
| 338 |
+
"rewards/rejected": -21.277738571166992,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9606986899563319,
|
| 343 |
+
"grad_norm": 0.007519610282738468,
|
| 344 |
+
"learning_rate": 2.351139701825267e-08,
|
| 345 |
+
"logits/chosen": -0.1052924394607544,
|
| 346 |
+
"logits/rejected": 1.4221456050872803,
|
| 347 |
+
"logps/chosen": -339.9599914550781,
|
| 348 |
+
"logps/rejected": -2215.8388671875,
|
| 349 |
+
"loss": 0.0,
|
| 350 |
+
"rewards/accuracies": 1.0,
|
| 351 |
+
"rewards/chosen": -0.8596421480178833,
|
| 352 |
+
"rewards/margins": 19.117597579956055,
|
| 353 |
+
"rewards/rejected": -19.97723960876465,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.0,
|
| 358 |
+
"step": 229,
|
| 359 |
+
"total_flos": 0.0,
|
| 360 |
+
"train_loss": 0.15411589154528807,
|
| 361 |
+
"train_runtime": 2209.5741,
|
| 362 |
+
"train_samples_per_second": 6.627,
|
| 363 |
+
"train_steps_per_second": 0.104
|
| 364 |
+
}
|
| 365 |
+
],
|
| 366 |
+
"logging_steps": 10,
|
| 367 |
+
"max_steps": 229,
|
| 368 |
+
"num_input_tokens_seen": 0,
|
| 369 |
+
"num_train_epochs": 1,
|
| 370 |
+
"save_steps": 100,
|
| 371 |
+
"stateful_callbacks": {
|
| 372 |
+
"TrainerControl": {
|
| 373 |
+
"args": {
|
| 374 |
+
"should_epoch_stop": false,
|
| 375 |
+
"should_evaluate": false,
|
| 376 |
+
"should_log": false,
|
| 377 |
+
"should_save": true,
|
| 378 |
+
"should_training_stop": true
|
| 379 |
+
},
|
| 380 |
+
"attributes": {}
|
| 381 |
+
}
|
| 382 |
+
},
|
| 383 |
+
"total_flos": 0.0,
|
| 384 |
+
"train_batch_size": 2,
|
| 385 |
+
"trial_name": null,
|
| 386 |
+
"trial_params": null
|
| 387 |
+
}
|