Commit
•
74b8803
1
Parent(s):
8fb11f5
Model save
Browse files- README.md +70 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- train_results.json +9 -0
- trainer_state.json +298 -0
README.md
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: AmberYifan/mistral-safe-sft-full
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: mistral-sft-kcenter-5k
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# mistral-sft-kcenter-5k
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [AmberYifan/mistral-safe-sft-full](https://huggingface.co/AmberYifan/mistral-safe-sft-full) on an unknown dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 1.0181
|
| 19 |
+
- Rewards/real: 1.4812
|
| 20 |
+
- Rewards/generated: 1.4157
|
| 21 |
+
- Rewards/accuracies: 0.0812
|
| 22 |
+
- Rewards/margins: 0.0655
|
| 23 |
+
- Logps/generated: -57.7408
|
| 24 |
+
- Logps/real: -16.8189
|
| 25 |
+
- Logits/generated: -2.6921
|
| 26 |
+
- Logits/real: -2.4190
|
| 27 |
+
|
| 28 |
+
## Model description
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Intended uses & limitations
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training and evaluation data
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Training procedure
|
| 41 |
+
|
| 42 |
+
### Training hyperparameters
|
| 43 |
+
|
| 44 |
+
The following hyperparameters were used during training:
|
| 45 |
+
- learning_rate: 5e-07
|
| 46 |
+
- train_batch_size: 8
|
| 47 |
+
- eval_batch_size: 8
|
| 48 |
+
- seed: 42
|
| 49 |
+
- distributed_type: multi-GPU
|
| 50 |
+
- num_devices: 4
|
| 51 |
+
- total_train_batch_size: 32
|
| 52 |
+
- total_eval_batch_size: 32
|
| 53 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 54 |
+
- lr_scheduler_type: linear
|
| 55 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 56 |
+
- num_epochs: 1
|
| 57 |
+
|
| 58 |
+
### Training results
|
| 59 |
+
|
| 60 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/real | Rewards/generated | Rewards/accuracies | Rewards/margins | Logps/generated | Logps/real | Logits/generated | Logits/real |
|
| 61 |
+
|:-------------:|:------:|:----:|:---------------:|:------------:|:-----------------:|:------------------:|:---------------:|:---------------:|:----------:|:----------------:|:-----------:|
|
| 62 |
+
| 0.3429 | 0.6369 | 100 | 1.0181 | 1.4812 | 1.4157 | 0.0812 | 0.0655 | -57.7408 | -16.8189 | -2.6921 | -2.4190 |
|
| 63 |
+
|
| 64 |
+
|
| 65 |
+
### Framework versions
|
| 66 |
+
|
| 67 |
+
- Transformers 4.43.3
|
| 68 |
+
- Pytorch 2.2.2+cu121
|
| 69 |
+
- Datasets 2.20.0
|
| 70 |
+
- 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": 0.3886139180250229,
|
| 5 |
+
"train_runtime": 1518.6555,
|
| 6 |
+
"train_samples": 5000,
|
| 7 |
+
"train_samples_per_second": 3.292,
|
| 8 |
+
"train_steps_per_second": 0.103
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.43.3"
|
| 6 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.3886139180250229,
|
| 5 |
+
"train_runtime": 1518.6555,
|
| 6 |
+
"train_samples": 5000,
|
| 7 |
+
"train_samples_per_second": 3.292,
|
| 8 |
+
"train_steps_per_second": 0.103
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,298 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 157,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.006369426751592357,
|
| 13 |
+
"grad_norm": 436.30790272840835,
|
| 14 |
+
"learning_rate": 3.125e-08,
|
| 15 |
+
"logits/generated": -2.574023723602295,
|
| 16 |
+
"logits/real": -2.507308006286621,
|
| 17 |
+
"logps/generated": -396.6226806640625,
|
| 18 |
+
"logps/real": -228.6820831298828,
|
| 19 |
+
"loss": 1.1036,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/generated": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/real": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.06369426751592357,
|
| 28 |
+
"grad_norm": 86.39288206954664,
|
| 29 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 30 |
+
"logits/generated": -2.6174476146698,
|
| 31 |
+
"logits/real": -2.577033042907715,
|
| 32 |
+
"logps/generated": -314.8096923828125,
|
| 33 |
+
"logps/real": -233.916259765625,
|
| 34 |
+
"loss": 0.8432,
|
| 35 |
+
"rewards/accuracies": 0.8055555820465088,
|
| 36 |
+
"rewards/generated": -1.1551631689071655,
|
| 37 |
+
"rewards/margins": 1.29329252243042,
|
| 38 |
+
"rewards/real": 0.13812927901744843,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.12738853503184713,
|
| 43 |
+
"grad_norm": 23.096946473336036,
|
| 44 |
+
"learning_rate": 4.858156028368794e-07,
|
| 45 |
+
"logits/generated": -2.934861421585083,
|
| 46 |
+
"logits/real": -2.8925399780273438,
|
| 47 |
+
"logps/generated": -313.32012939453125,
|
| 48 |
+
"logps/real": -235.418701171875,
|
| 49 |
+
"loss": 0.4872,
|
| 50 |
+
"rewards/accuracies": 1.0,
|
| 51 |
+
"rewards/generated": -3.896754741668701,
|
| 52 |
+
"rewards/margins": 6.176419258117676,
|
| 53 |
+
"rewards/real": 2.2796645164489746,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.1910828025477707,
|
| 58 |
+
"grad_norm": 23.454945687282052,
|
| 59 |
+
"learning_rate": 4.50354609929078e-07,
|
| 60 |
+
"logits/generated": -2.9848690032958984,
|
| 61 |
+
"logits/real": -2.9865798950195312,
|
| 62 |
+
"logps/generated": -304.69854736328125,
|
| 63 |
+
"logps/real": -152.80667114257812,
|
| 64 |
+
"loss": 0.3868,
|
| 65 |
+
"rewards/accuracies": 0.987500011920929,
|
| 66 |
+
"rewards/generated": -3.0515201091766357,
|
| 67 |
+
"rewards/margins": 7.164666652679443,
|
| 68 |
+
"rewards/real": 4.113146781921387,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.25477707006369427,
|
| 73 |
+
"grad_norm": 56.59285724233864,
|
| 74 |
+
"learning_rate": 4.148936170212766e-07,
|
| 75 |
+
"logits/generated": -2.821133852005005,
|
| 76 |
+
"logits/real": -2.894505262374878,
|
| 77 |
+
"logps/generated": -349.6683044433594,
|
| 78 |
+
"logps/real": -201.2012176513672,
|
| 79 |
+
"loss": 0.4438,
|
| 80 |
+
"rewards/accuracies": 1.0,
|
| 81 |
+
"rewards/generated": -5.3700079917907715,
|
| 82 |
+
"rewards/margins": 9.444425582885742,
|
| 83 |
+
"rewards/real": 4.074416637420654,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.3184713375796178,
|
| 88 |
+
"grad_norm": 26.344052963980364,
|
| 89 |
+
"learning_rate": 3.7943262411347514e-07,
|
| 90 |
+
"logits/generated": -3.041593074798584,
|
| 91 |
+
"logits/real": -2.9904277324676514,
|
| 92 |
+
"logps/generated": -335.09942626953125,
|
| 93 |
+
"logps/real": -158.0336151123047,
|
| 94 |
+
"loss": 0.37,
|
| 95 |
+
"rewards/accuracies": 0.987500011920929,
|
| 96 |
+
"rewards/generated": -3.180644989013672,
|
| 97 |
+
"rewards/margins": 8.93952751159668,
|
| 98 |
+
"rewards/real": 5.758882999420166,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.3821656050955414,
|
| 103 |
+
"grad_norm": 47.871426260084256,
|
| 104 |
+
"learning_rate": 3.4397163120567375e-07,
|
| 105 |
+
"logits/generated": -2.9896843433380127,
|
| 106 |
+
"logits/real": -3.0180420875549316,
|
| 107 |
+
"logps/generated": -302.9552307128906,
|
| 108 |
+
"logps/real": -140.7099151611328,
|
| 109 |
+
"loss": 0.3324,
|
| 110 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 111 |
+
"rewards/generated": -3.2430217266082764,
|
| 112 |
+
"rewards/margins": 9.301046371459961,
|
| 113 |
+
"rewards/real": 6.0580244064331055,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.445859872611465,
|
| 118 |
+
"grad_norm": 24.288240218254995,
|
| 119 |
+
"learning_rate": 3.085106382978723e-07,
|
| 120 |
+
"logits/generated": -3.0595779418945312,
|
| 121 |
+
"logits/real": -2.9718518257141113,
|
| 122 |
+
"logps/generated": -328.691650390625,
|
| 123 |
+
"logps/real": -140.92660522460938,
|
| 124 |
+
"loss": 0.333,
|
| 125 |
+
"rewards/accuracies": 0.987500011920929,
|
| 126 |
+
"rewards/generated": -3.642324447631836,
|
| 127 |
+
"rewards/margins": 10.118036270141602,
|
| 128 |
+
"rewards/real": 6.47571325302124,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5095541401273885,
|
| 133 |
+
"grad_norm": 14.83120429808784,
|
| 134 |
+
"learning_rate": 2.730496453900709e-07,
|
| 135 |
+
"logits/generated": -2.996634006500244,
|
| 136 |
+
"logits/real": -3.0401620864868164,
|
| 137 |
+
"logps/generated": -322.6135559082031,
|
| 138 |
+
"logps/real": -159.5257110595703,
|
| 139 |
+
"loss": 0.3328,
|
| 140 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 141 |
+
"rewards/generated": -2.347047805786133,
|
| 142 |
+
"rewards/margins": 10.133938789367676,
|
| 143 |
+
"rewards/real": 7.786890506744385,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.5732484076433121,
|
| 148 |
+
"grad_norm": 16.65278970025503,
|
| 149 |
+
"learning_rate": 2.375886524822695e-07,
|
| 150 |
+
"logits/generated": -3.0264060497283936,
|
| 151 |
+
"logits/real": -2.979038715362549,
|
| 152 |
+
"logps/generated": -360.536376953125,
|
| 153 |
+
"logps/real": -162.26622009277344,
|
| 154 |
+
"loss": 0.3387,
|
| 155 |
+
"rewards/accuracies": 0.987500011920929,
|
| 156 |
+
"rewards/generated": -4.7210164070129395,
|
| 157 |
+
"rewards/margins": 12.409219741821289,
|
| 158 |
+
"rewards/real": 7.68820333480835,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.6369426751592356,
|
| 163 |
+
"grad_norm": 7.324308799399964,
|
| 164 |
+
"learning_rate": 2.0212765957446807e-07,
|
| 165 |
+
"logits/generated": -2.968139410018921,
|
| 166 |
+
"logits/real": -3.0011098384857178,
|
| 167 |
+
"logps/generated": -325.79107666015625,
|
| 168 |
+
"logps/real": -162.4396514892578,
|
| 169 |
+
"loss": 0.3429,
|
| 170 |
+
"rewards/accuracies": 0.987500011920929,
|
| 171 |
+
"rewards/generated": -2.580188512802124,
|
| 172 |
+
"rewards/margins": 10.546457290649414,
|
| 173 |
+
"rewards/real": 7.966268062591553,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.6369426751592356,
|
| 178 |
+
"eval_logits/generated": -2.6921370029449463,
|
| 179 |
+
"eval_logits/real": -2.418954372406006,
|
| 180 |
+
"eval_logps/generated": -57.74082565307617,
|
| 181 |
+
"eval_logps/real": -16.818851470947266,
|
| 182 |
+
"eval_loss": 1.018080234527588,
|
| 183 |
+
"eval_rewards/accuracies": 0.0812101885676384,
|
| 184 |
+
"eval_rewards/generated": 1.4157419204711914,
|
| 185 |
+
"eval_rewards/margins": 0.06545643508434296,
|
| 186 |
+
"eval_rewards/real": 1.4811984300613403,
|
| 187 |
+
"eval_runtime": 325.124,
|
| 188 |
+
"eval_samples_per_second": 15.379,
|
| 189 |
+
"eval_steps_per_second": 0.483,
|
| 190 |
+
"step": 100
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.7006369426751592,
|
| 194 |
+
"grad_norm": 44.79401383111945,
|
| 195 |
+
"learning_rate": 1.6666666666666665e-07,
|
| 196 |
+
"logits/generated": -2.966001033782959,
|
| 197 |
+
"logits/real": -2.9780120849609375,
|
| 198 |
+
"logps/generated": -347.74993896484375,
|
| 199 |
+
"logps/real": -158.44979858398438,
|
| 200 |
+
"loss": 0.365,
|
| 201 |
+
"rewards/accuracies": 0.987500011920929,
|
| 202 |
+
"rewards/generated": -3.206289768218994,
|
| 203 |
+
"rewards/margins": 11.313484191894531,
|
| 204 |
+
"rewards/real": 8.107194900512695,
|
| 205 |
+
"step": 110
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7643312101910829,
|
| 209 |
+
"grad_norm": 7.15542526632935,
|
| 210 |
+
"learning_rate": 1.3120567375886523e-07,
|
| 211 |
+
"logits/generated": -2.9361348152160645,
|
| 212 |
+
"logits/real": -2.991243839263916,
|
| 213 |
+
"logps/generated": -340.3627014160156,
|
| 214 |
+
"logps/real": -145.40615844726562,
|
| 215 |
+
"loss": 0.3243,
|
| 216 |
+
"rewards/accuracies": 1.0,
|
| 217 |
+
"rewards/generated": -3.4838199615478516,
|
| 218 |
+
"rewards/margins": 10.616706848144531,
|
| 219 |
+
"rewards/real": 7.132887363433838,
|
| 220 |
+
"step": 120
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.8280254777070064,
|
| 224 |
+
"grad_norm": 28.617693401838523,
|
| 225 |
+
"learning_rate": 9.574468085106382e-08,
|
| 226 |
+
"logits/generated": -2.880622386932373,
|
| 227 |
+
"logits/real": -2.873990774154663,
|
| 228 |
+
"logps/generated": -284.91949462890625,
|
| 229 |
+
"logps/real": -130.95306396484375,
|
| 230 |
+
"loss": 0.3022,
|
| 231 |
+
"rewards/accuracies": 0.987500011920929,
|
| 232 |
+
"rewards/generated": -1.7718982696533203,
|
| 233 |
+
"rewards/margins": 8.987956047058105,
|
| 234 |
+
"rewards/real": 7.216057777404785,
|
| 235 |
+
"step": 130
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.89171974522293,
|
| 239 |
+
"grad_norm": 11.2855453725489,
|
| 240 |
+
"learning_rate": 6.02836879432624e-08,
|
| 241 |
+
"logits/generated": -2.893979549407959,
|
| 242 |
+
"logits/real": -2.919368028640747,
|
| 243 |
+
"logps/generated": -334.3547058105469,
|
| 244 |
+
"logps/real": -154.1964569091797,
|
| 245 |
+
"loss": 0.3389,
|
| 246 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 247 |
+
"rewards/generated": -3.628894090652466,
|
| 248 |
+
"rewards/margins": 11.9329252243042,
|
| 249 |
+
"rewards/real": 8.304032325744629,
|
| 250 |
+
"step": 140
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.9554140127388535,
|
| 254 |
+
"grad_norm": 31.41907953329688,
|
| 255 |
+
"learning_rate": 2.4822695035460993e-08,
|
| 256 |
+
"logits/generated": -2.864701747894287,
|
| 257 |
+
"logits/real": -2.9138083457946777,
|
| 258 |
+
"logps/generated": -315.903564453125,
|
| 259 |
+
"logps/real": -138.75643920898438,
|
| 260 |
+
"loss": 0.3255,
|
| 261 |
+
"rewards/accuracies": 0.987500011920929,
|
| 262 |
+
"rewards/generated": -2.897620677947998,
|
| 263 |
+
"rewards/margins": 9.965813636779785,
|
| 264 |
+
"rewards/real": 7.068192958831787,
|
| 265 |
+
"step": 150
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.0,
|
| 269 |
+
"step": 157,
|
| 270 |
+
"total_flos": 0.0,
|
| 271 |
+
"train_loss": 0.3886139180250229,
|
| 272 |
+
"train_runtime": 1518.6555,
|
| 273 |
+
"train_samples_per_second": 3.292,
|
| 274 |
+
"train_steps_per_second": 0.103
|
| 275 |
+
}
|
| 276 |
+
],
|
| 277 |
+
"logging_steps": 10,
|
| 278 |
+
"max_steps": 157,
|
| 279 |
+
"num_input_tokens_seen": 0,
|
| 280 |
+
"num_train_epochs": 1,
|
| 281 |
+
"save_steps": 100,
|
| 282 |
+
"stateful_callbacks": {
|
| 283 |
+
"TrainerControl": {
|
| 284 |
+
"args": {
|
| 285 |
+
"should_epoch_stop": false,
|
| 286 |
+
"should_evaluate": false,
|
| 287 |
+
"should_log": false,
|
| 288 |
+
"should_save": true,
|
| 289 |
+
"should_training_stop": true
|
| 290 |
+
},
|
| 291 |
+
"attributes": {}
|
| 292 |
+
}
|
| 293 |
+
},
|
| 294 |
+
"total_flos": 0.0,
|
| 295 |
+
"train_batch_size": 8,
|
| 296 |
+
"trial_name": null,
|
| 297 |
+
"trial_params": null
|
| 298 |
+
}
|