Model save
Browse files- README.md +61 -0
- all_results.json +9 -0
- generation_config.json +7 -0
- train_results.json +9 -0
- trainer_state.json +357 -0
README.md
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
|
| 5 |
+
tags:
|
| 6 |
+
- trl
|
| 7 |
+
- dpo
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
model-index:
|
| 10 |
+
- name: tinyllama-1.1b-chat-v1.0-ui-math-coding-dpo
|
| 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 |
+
# tinyllama-1.1b-chat-v1.0-ui-math-coding-dpo
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) 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-07
|
| 39 |
+
- train_batch_size: 16
|
| 40 |
+
- eval_batch_size: 16
|
| 41 |
+
- seed: 42
|
| 42 |
+
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 8
|
| 44 |
+
- gradient_accumulation_steps: 8
|
| 45 |
+
- total_train_batch_size: 1024
|
| 46 |
+
- total_eval_batch_size: 128
|
| 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 |
+
- Transformers 4.44.1
|
| 59 |
+
- Pytorch 2.1.2+cu121
|
| 60 |
+
- Datasets 2.21.0
|
| 61 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9963369963369964,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.4527332771058176,
|
| 5 |
+
"train_runtime": 5381.1516,
|
| 6 |
+
"train_samples": 209650,
|
| 7 |
+
"train_samples_per_second": 38.96,
|
| 8 |
+
"train_steps_per_second": 0.038
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"eos_token_id": 2,
|
| 4 |
+
"max_length": 2048,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.44.1"
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9963369963369964,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.4527332771058176,
|
| 5 |
+
"train_runtime": 5381.1516,
|
| 6 |
+
"train_samples": 209650,
|
| 7 |
+
"train_samples_per_second": 38.96,
|
| 8 |
+
"train_steps_per_second": 0.038
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,357 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9963369963369964,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 204,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004884004884004884,
|
| 13 |
+
"grad_norm": 2.3243459220572165,
|
| 14 |
+
"learning_rate": 2.3809523809523807e-08,
|
| 15 |
+
"logits/chosen": -2.550273895263672,
|
| 16 |
+
"logits/rejected": -2.5806894302368164,
|
| 17 |
+
"logps/chosen": -424.7008056640625,
|
| 18 |
+
"logps/rejected": -390.49554443359375,
|
| 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": 0.04884004884004884,
|
| 28 |
+
"grad_norm": 2.4712584301903604,
|
| 29 |
+
"learning_rate": 2.3809523809523806e-07,
|
| 30 |
+
"logits/chosen": -2.4481005668640137,
|
| 31 |
+
"logits/rejected": -2.474926471710205,
|
| 32 |
+
"logps/chosen": -395.8595886230469,
|
| 33 |
+
"logps/rejected": -384.5038146972656,
|
| 34 |
+
"loss": 0.6931,
|
| 35 |
+
"rewards/accuracies": 0.4635416567325592,
|
| 36 |
+
"rewards/chosen": 0.00013807932555209845,
|
| 37 |
+
"rewards/margins": 0.0004333473916631192,
|
| 38 |
+
"rewards/rejected": -0.00029526810976676643,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.09768009768009768,
|
| 43 |
+
"grad_norm": 2.318089253747947,
|
| 44 |
+
"learning_rate": 4.761904761904761e-07,
|
| 45 |
+
"logits/chosen": -2.445664167404175,
|
| 46 |
+
"logits/rejected": -2.4723546504974365,
|
| 47 |
+
"logps/chosen": -393.4665222167969,
|
| 48 |
+
"logps/rejected": -377.8502197265625,
|
| 49 |
+
"loss": 0.6905,
|
| 50 |
+
"rewards/accuracies": 0.67578125,
|
| 51 |
+
"rewards/chosen": 0.00608012406155467,
|
| 52 |
+
"rewards/margins": 0.005808630492538214,
|
| 53 |
+
"rewards/rejected": 0.00027149339439347386,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.14652014652014653,
|
| 58 |
+
"grad_norm": 2.5283415680520225,
|
| 59 |
+
"learning_rate": 4.970219740227693e-07,
|
| 60 |
+
"logits/chosen": -2.494197368621826,
|
| 61 |
+
"logits/rejected": -2.5383658409118652,
|
| 62 |
+
"logps/chosen": -393.2831115722656,
|
| 63 |
+
"logps/rejected": -383.0456237792969,
|
| 64 |
+
"loss": 0.6754,
|
| 65 |
+
"rewards/accuracies": 0.807812511920929,
|
| 66 |
+
"rewards/chosen": 0.03202961012721062,
|
| 67 |
+
"rewards/margins": 0.037289537489414215,
|
| 68 |
+
"rewards/rejected": -0.005259926896542311,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.19536019536019536,
|
| 73 |
+
"grad_norm": 2.5364879915405267,
|
| 74 |
+
"learning_rate": 4.868186180746791e-07,
|
| 75 |
+
"logits/chosen": -2.5153324604034424,
|
| 76 |
+
"logits/rejected": -2.5360398292541504,
|
| 77 |
+
"logps/chosen": -387.1582946777344,
|
| 78 |
+
"logps/rejected": -379.3692932128906,
|
| 79 |
+
"loss": 0.647,
|
| 80 |
+
"rewards/accuracies": 0.8218749761581421,
|
| 81 |
+
"rewards/chosen": 0.0628650039434433,
|
| 82 |
+
"rewards/margins": 0.09538714587688446,
|
| 83 |
+
"rewards/rejected": -0.032522134482860565,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.2442002442002442,
|
| 88 |
+
"grad_norm": 2.2699251495146964,
|
| 89 |
+
"learning_rate": 4.6965306126428705e-07,
|
| 90 |
+
"logits/chosen": -2.539130449295044,
|
| 91 |
+
"logits/rejected": -2.5619969367980957,
|
| 92 |
+
"logps/chosen": -404.4756774902344,
|
| 93 |
+
"logps/rejected": -406.6902770996094,
|
| 94 |
+
"loss": 0.5925,
|
| 95 |
+
"rewards/accuracies": 0.82421875,
|
| 96 |
+
"rewards/chosen": -0.0003643702657427639,
|
| 97 |
+
"rewards/margins": 0.2586084008216858,
|
| 98 |
+
"rewards/rejected": -0.25897279381752014,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.29304029304029305,
|
| 103 |
+
"grad_norm": 2.4145672311111994,
|
| 104 |
+
"learning_rate": 4.460299516441776e-07,
|
| 105 |
+
"logits/chosen": -2.550515651702881,
|
| 106 |
+
"logits/rejected": -2.577197551727295,
|
| 107 |
+
"logps/chosen": -413.14947509765625,
|
| 108 |
+
"logps/rejected": -442.47674560546875,
|
| 109 |
+
"loss": 0.5417,
|
| 110 |
+
"rewards/accuracies": 0.801562488079071,
|
| 111 |
+
"rewards/chosen": -0.15052883327007294,
|
| 112 |
+
"rewards/margins": 0.4667808413505554,
|
| 113 |
+
"rewards/rejected": -0.6173096895217896,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3418803418803419,
|
| 118 |
+
"grad_norm": 2.7821486929434056,
|
| 119 |
+
"learning_rate": 4.166437820523908e-07,
|
| 120 |
+
"logits/chosen": -2.5242340564727783,
|
| 121 |
+
"logits/rejected": -2.5425312519073486,
|
| 122 |
+
"logps/chosen": -446.6337890625,
|
| 123 |
+
"logps/rejected": -487.9483337402344,
|
| 124 |
+
"loss": 0.5011,
|
| 125 |
+
"rewards/accuracies": 0.8023437261581421,
|
| 126 |
+
"rewards/chosen": -0.4636126160621643,
|
| 127 |
+
"rewards/margins": 0.6593486070632935,
|
| 128 |
+
"rewards/rejected": -1.1229612827301025,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.3907203907203907,
|
| 133 |
+
"grad_norm": 2.754390105851781,
|
| 134 |
+
"learning_rate": 3.8235847280454626e-07,
|
| 135 |
+
"logits/chosen": -2.4175186157226562,
|
| 136 |
+
"logits/rejected": -2.449018955230713,
|
| 137 |
+
"logps/chosen": -507.10357666015625,
|
| 138 |
+
"logps/rejected": -592.0707397460938,
|
| 139 |
+
"loss": 0.4575,
|
| 140 |
+
"rewards/accuracies": 0.80078125,
|
| 141 |
+
"rewards/chosen": -1.1133525371551514,
|
| 142 |
+
"rewards/margins": 0.9770663380622864,
|
| 143 |
+
"rewards/rejected": -2.090418815612793,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.43956043956043955,
|
| 148 |
+
"grad_norm": 2.6724675179161568,
|
| 149 |
+
"learning_rate": 3.4418197340879627e-07,
|
| 150 |
+
"logits/chosen": -2.409747838973999,
|
| 151 |
+
"logits/rejected": -2.4172959327697754,
|
| 152 |
+
"logps/chosen": -507.91876220703125,
|
| 153 |
+
"logps/rejected": -625.10205078125,
|
| 154 |
+
"loss": 0.4121,
|
| 155 |
+
"rewards/accuracies": 0.835156261920929,
|
| 156 |
+
"rewards/chosen": -1.0922380685806274,
|
| 157 |
+
"rewards/margins": 1.3621976375579834,
|
| 158 |
+
"rewards/rejected": -2.4544358253479004,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.4884004884004884,
|
| 163 |
+
"grad_norm": 3.151335337005712,
|
| 164 |
+
"learning_rate": 3.032366299846039e-07,
|
| 165 |
+
"logits/chosen": -2.4340109825134277,
|
| 166 |
+
"logits/rejected": -2.4465105533599854,
|
| 167 |
+
"logps/chosen": -532.0742797851562,
|
| 168 |
+
"logps/rejected": -683.5274658203125,
|
| 169 |
+
"loss": 0.3892,
|
| 170 |
+
"rewards/accuracies": 0.8335937261581421,
|
| 171 |
+
"rewards/chosen": -1.2753849029541016,
|
| 172 |
+
"rewards/margins": 1.6482696533203125,
|
| 173 |
+
"rewards/rejected": -2.923654794692993,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.5372405372405372,
|
| 178 |
+
"grad_norm": 2.717335654672678,
|
| 179 |
+
"learning_rate": 2.6072618954988863e-07,
|
| 180 |
+
"logits/chosen": -2.4394848346710205,
|
| 181 |
+
"logits/rejected": -2.442568778991699,
|
| 182 |
+
"logps/chosen": -518.7210693359375,
|
| 183 |
+
"logps/rejected": -677.293701171875,
|
| 184 |
+
"loss": 0.3834,
|
| 185 |
+
"rewards/accuracies": 0.827343761920929,
|
| 186 |
+
"rewards/chosen": -1.2031551599502563,
|
| 187 |
+
"rewards/margins": 1.6999378204345703,
|
| 188 |
+
"rewards/rejected": -2.903092861175537,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.5860805860805861,
|
| 193 |
+
"grad_norm": 3.1603927594667005,
|
| 194 |
+
"learning_rate": 2.1790041121336222e-07,
|
| 195 |
+
"logits/chosen": -2.4521875381469727,
|
| 196 |
+
"logits/rejected": -2.460845470428467,
|
| 197 |
+
"logps/chosen": -532.1248168945312,
|
| 198 |
+
"logps/rejected": -704.0490112304688,
|
| 199 |
+
"loss": 0.3613,
|
| 200 |
+
"rewards/accuracies": 0.8414062261581421,
|
| 201 |
+
"rewards/chosen": -1.3992774486541748,
|
| 202 |
+
"rewards/margins": 1.8053524494171143,
|
| 203 |
+
"rewards/rejected": -3.204629898071289,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.6349206349206349,
|
| 208 |
+
"grad_norm": 2.877029930356179,
|
| 209 |
+
"learning_rate": 1.7601832466317766e-07,
|
| 210 |
+
"logits/chosen": -2.4438443183898926,
|
| 211 |
+
"logits/rejected": -2.462118148803711,
|
| 212 |
+
"logps/chosen": -540.3773193359375,
|
| 213 |
+
"logps/rejected": -711.51416015625,
|
| 214 |
+
"loss": 0.3581,
|
| 215 |
+
"rewards/accuracies": 0.8453124761581421,
|
| 216 |
+
"rewards/chosen": -1.4280272722244263,
|
| 217 |
+
"rewards/margins": 1.8352330923080444,
|
| 218 |
+
"rewards/rejected": -3.2632603645324707,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6837606837606838,
|
| 223 |
+
"grad_norm": 2.8365028089984454,
|
| 224 |
+
"learning_rate": 1.3631121611097362e-07,
|
| 225 |
+
"logits/chosen": -2.4740078449249268,
|
| 226 |
+
"logits/rejected": -2.487417697906494,
|
| 227 |
+
"logps/chosen": -546.05859375,
|
| 228 |
+
"logps/rejected": -727.7886962890625,
|
| 229 |
+
"loss": 0.3495,
|
| 230 |
+
"rewards/accuracies": 0.8609374761581421,
|
| 231 |
+
"rewards/chosen": -1.4827759265899658,
|
| 232 |
+
"rewards/margins": 1.9239017963409424,
|
| 233 |
+
"rewards/rejected": -3.4066779613494873,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.7326007326007326,
|
| 238 |
+
"grad_norm": 3.083741716442478,
|
| 239 |
+
"learning_rate": 9.9946429862908e-08,
|
| 240 |
+
"logits/chosen": -2.462756633758545,
|
| 241 |
+
"logits/rejected": -2.4654526710510254,
|
| 242 |
+
"logps/chosen": -549.0475463867188,
|
| 243 |
+
"logps/rejected": -722.2012329101562,
|
| 244 |
+
"loss": 0.3454,
|
| 245 |
+
"rewards/accuracies": 0.8492187261581421,
|
| 246 |
+
"rewards/chosen": -1.5003674030303955,
|
| 247 |
+
"rewards/margins": 1.9490848779678345,
|
| 248 |
+
"rewards/rejected": -3.4494519233703613,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.7814407814407814,
|
| 253 |
+
"grad_norm": 3.086204461780561,
|
| 254 |
+
"learning_rate": 6.799304971075381e-08,
|
| 255 |
+
"logits/chosen": -2.4620633125305176,
|
| 256 |
+
"logits/rejected": -2.4670565128326416,
|
| 257 |
+
"logps/chosen": -539.4750366210938,
|
| 258 |
+
"logps/rejected": -717.6760864257812,
|
| 259 |
+
"loss": 0.3426,
|
| 260 |
+
"rewards/accuracies": 0.859375,
|
| 261 |
+
"rewards/chosen": -1.4655063152313232,
|
| 262 |
+
"rewards/margins": 1.9486806392669678,
|
| 263 |
+
"rewards/rejected": -3.41418719291687,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.8302808302808303,
|
| 268 |
+
"grad_norm": 3.152043777770028,
|
| 269 |
+
"learning_rate": 4.1390469071538175e-08,
|
| 270 |
+
"logits/chosen": -2.4839229583740234,
|
| 271 |
+
"logits/rejected": -2.4979355335235596,
|
| 272 |
+
"logps/chosen": -547.6788940429688,
|
| 273 |
+
"logps/rejected": -723.7260131835938,
|
| 274 |
+
"loss": 0.3417,
|
| 275 |
+
"rewards/accuracies": 0.839062511920929,
|
| 276 |
+
"rewards/chosen": -1.532496690750122,
|
| 277 |
+
"rewards/margins": 1.8789927959442139,
|
| 278 |
+
"rewards/rejected": -3.411489486694336,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.8791208791208791,
|
| 283 |
+
"grad_norm": 3.34422267800285,
|
| 284 |
+
"learning_rate": 2.0920773878248837e-08,
|
| 285 |
+
"logits/chosen": -2.4812464714050293,
|
| 286 |
+
"logits/rejected": -2.49059796333313,
|
| 287 |
+
"logps/chosen": -556.6783447265625,
|
| 288 |
+
"logps/rejected": -743.0768432617188,
|
| 289 |
+
"loss": 0.3413,
|
| 290 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 291 |
+
"rewards/chosen": -1.5343992710113525,
|
| 292 |
+
"rewards/margins": 2.0101265907287598,
|
| 293 |
+
"rewards/rejected": -3.5445258617401123,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.927960927960928,
|
| 298 |
+
"grad_norm": 3.2655151670502574,
|
| 299 |
+
"learning_rate": 7.185750133542168e-09,
|
| 300 |
+
"logits/chosen": -2.473402500152588,
|
| 301 |
+
"logits/rejected": -2.471391201019287,
|
| 302 |
+
"logps/chosen": -554.5660400390625,
|
| 303 |
+
"logps/rejected": -741.9495849609375,
|
| 304 |
+
"loss": 0.3343,
|
| 305 |
+
"rewards/accuracies": 0.8539062738418579,
|
| 306 |
+
"rewards/chosen": -1.5347990989685059,
|
| 307 |
+
"rewards/margins": 2.0656068325042725,
|
| 308 |
+
"rewards/rejected": -3.6004059314727783,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.9768009768009768,
|
| 313 |
+
"grad_norm": 4.051209744645471,
|
| 314 |
+
"learning_rate": 5.891920784984184e-10,
|
| 315 |
+
"logits/chosen": -2.456406354904175,
|
| 316 |
+
"logits/rejected": -2.4556210041046143,
|
| 317 |
+
"logps/chosen": -550.1729125976562,
|
| 318 |
+
"logps/rejected": -737.451904296875,
|
| 319 |
+
"loss": 0.3391,
|
| 320 |
+
"rewards/accuracies": 0.875,
|
| 321 |
+
"rewards/chosen": -1.517017126083374,
|
| 322 |
+
"rewards/margins": 2.044365406036377,
|
| 323 |
+
"rewards/rejected": -3.561382293701172,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.9963369963369964,
|
| 328 |
+
"step": 204,
|
| 329 |
+
"total_flos": 0.0,
|
| 330 |
+
"train_loss": 0.4527332771058176,
|
| 331 |
+
"train_runtime": 5381.1516,
|
| 332 |
+
"train_samples_per_second": 38.96,
|
| 333 |
+
"train_steps_per_second": 0.038
|
| 334 |
+
}
|
| 335 |
+
],
|
| 336 |
+
"logging_steps": 10,
|
| 337 |
+
"max_steps": 204,
|
| 338 |
+
"num_input_tokens_seen": 0,
|
| 339 |
+
"num_train_epochs": 1,
|
| 340 |
+
"save_steps": 100,
|
| 341 |
+
"stateful_callbacks": {
|
| 342 |
+
"TrainerControl": {
|
| 343 |
+
"args": {
|
| 344 |
+
"should_epoch_stop": false,
|
| 345 |
+
"should_evaluate": false,
|
| 346 |
+
"should_log": false,
|
| 347 |
+
"should_save": true,
|
| 348 |
+
"should_training_stop": true
|
| 349 |
+
},
|
| 350 |
+
"attributes": {}
|
| 351 |
+
}
|
| 352 |
+
},
|
| 353 |
+
"total_flos": 0.0,
|
| 354 |
+
"train_batch_size": 16,
|
| 355 |
+
"trial_name": null,
|
| 356 |
+
"trial_params": null
|
| 357 |
+
}
|