Model save
Browse files- README.md +79 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +216 -0
README.md
ADDED
|
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: mistralai/Mistral-7B-Instruct-v0.1
|
| 3 |
+
datasets:
|
| 4 |
+
- generator
|
| 5 |
+
library_name: peft
|
| 6 |
+
license: apache-2.0
|
| 7 |
+
tags:
|
| 8 |
+
- trl
|
| 9 |
+
- sft
|
| 10 |
+
- generated_from_trainer
|
| 11 |
+
model-index:
|
| 12 |
+
- name: mistral_cot_true_simple_lora
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# mistral_cot_true_simple_lora
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.9395
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0002
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
+
- eval_batch_size: 4
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- gradient_accumulation_steps: 2
|
| 48 |
+
- total_train_batch_size: 8
|
| 49 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
+
- lr_scheduler_type: cosine
|
| 51 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 52 |
+
- num_epochs: 15
|
| 53 |
+
|
| 54 |
+
### Training results
|
| 55 |
+
|
| 56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 57 |
+
|:-------------:|:-------:|:----:|:---------------:|
|
| 58 |
+
| 1.0104 | 0.8571 | 3 | 0.8812 |
|
| 59 |
+
| 0.8795 | 2.0 | 7 | 0.7078 |
|
| 60 |
+
| 0.6536 | 2.8571 | 10 | 0.6557 |
|
| 61 |
+
| 0.6536 | 4.0 | 14 | 0.6372 |
|
| 62 |
+
| 0.472 | 4.8571 | 17 | 0.6561 |
|
| 63 |
+
| 0.3263 | 6.0 | 21 | 0.6875 |
|
| 64 |
+
| 0.3263 | 6.8571 | 24 | 0.7510 |
|
| 65 |
+
| 0.1683 | 8.0 | 28 | 0.8222 |
|
| 66 |
+
| 0.0699 | 8.8571 | 31 | 0.8685 |
|
| 67 |
+
| 0.0369 | 10.0 | 35 | 0.9062 |
|
| 68 |
+
| 0.0369 | 10.8571 | 38 | 0.9268 |
|
| 69 |
+
| 0.0248 | 12.0 | 42 | 0.9377 |
|
| 70 |
+
| 0.0212 | 12.8571 | 45 | 0.9395 |
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
### Framework versions
|
| 74 |
+
|
| 75 |
+
- PEFT 0.12.0
|
| 76 |
+
- Transformers 4.44.2
|
| 77 |
+
- Pytorch 2.4.1+cu121
|
| 78 |
+
- Datasets 3.0.0
|
| 79 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.857142857142858,
|
| 3 |
+
"total_flos": 10453480636416.0,
|
| 4 |
+
"train_loss": 0.2976339568694433,
|
| 5 |
+
"train_runtime": 339.6314,
|
| 6 |
+
"train_samples": 100,
|
| 7 |
+
"train_samples_per_second": 1.148,
|
| 8 |
+
"train_steps_per_second": 0.132
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 12.857142857142858,
|
| 3 |
+
"total_flos": 10453480636416.0,
|
| 4 |
+
"train_loss": 0.2976339568694433,
|
| 5 |
+
"train_runtime": 339.6314,
|
| 6 |
+
"train_samples": 100,
|
| 7 |
+
"train_samples_per_second": 1.148,
|
| 8 |
+
"train_steps_per_second": 0.132
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 12.857142857142858,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 45,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2857142857142857,
|
| 13 |
+
"grad_norm": 0.7675084033789445,
|
| 14 |
+
"learning_rate": 4e-05,
|
| 15 |
+
"loss": 1.0104,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.8571428571428571,
|
| 20 |
+
"eval_loss": 0.8811686635017395,
|
| 21 |
+
"eval_runtime": 2.3203,
|
| 22 |
+
"eval_samples_per_second": 2.155,
|
| 23 |
+
"eval_steps_per_second": 0.862,
|
| 24 |
+
"step": 3
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.4285714285714286,
|
| 28 |
+
"grad_norm": 0.3794777121113468,
|
| 29 |
+
"learning_rate": 0.0002,
|
| 30 |
+
"loss": 0.8795,
|
| 31 |
+
"step": 5
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"eval_loss": 0.7077744007110596,
|
| 36 |
+
"eval_runtime": 1.1217,
|
| 37 |
+
"eval_samples_per_second": 4.458,
|
| 38 |
+
"eval_steps_per_second": 1.783,
|
| 39 |
+
"step": 7
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 2.857142857142857,
|
| 43 |
+
"grad_norm": 0.3530496345436479,
|
| 44 |
+
"learning_rate": 0.0001923879532511287,
|
| 45 |
+
"loss": 0.6536,
|
| 46 |
+
"step": 10
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 2.857142857142857,
|
| 50 |
+
"eval_loss": 0.6557201743125916,
|
| 51 |
+
"eval_runtime": 1.1188,
|
| 52 |
+
"eval_samples_per_second": 4.469,
|
| 53 |
+
"eval_steps_per_second": 1.788,
|
| 54 |
+
"step": 10
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 4.0,
|
| 58 |
+
"eval_loss": 0.6372172236442566,
|
| 59 |
+
"eval_runtime": 1.1233,
|
| 60 |
+
"eval_samples_per_second": 4.451,
|
| 61 |
+
"eval_steps_per_second": 1.781,
|
| 62 |
+
"step": 14
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.285714285714286,
|
| 66 |
+
"grad_norm": 0.26413927917508156,
|
| 67 |
+
"learning_rate": 0.00017071067811865476,
|
| 68 |
+
"loss": 0.472,
|
| 69 |
+
"step": 15
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 4.857142857142857,
|
| 73 |
+
"eval_loss": 0.6561022400856018,
|
| 74 |
+
"eval_runtime": 1.1268,
|
| 75 |
+
"eval_samples_per_second": 4.437,
|
| 76 |
+
"eval_steps_per_second": 1.775,
|
| 77 |
+
"step": 17
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.714285714285714,
|
| 81 |
+
"grad_norm": 0.3618434995437738,
|
| 82 |
+
"learning_rate": 0.000138268343236509,
|
| 83 |
+
"loss": 0.3263,
|
| 84 |
+
"step": 20
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 6.0,
|
| 88 |
+
"eval_loss": 0.687515914440155,
|
| 89 |
+
"eval_runtime": 1.1222,
|
| 90 |
+
"eval_samples_per_second": 4.456,
|
| 91 |
+
"eval_steps_per_second": 1.782,
|
| 92 |
+
"step": 21
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 6.857142857142857,
|
| 96 |
+
"eval_loss": 0.7509881854057312,
|
| 97 |
+
"eval_runtime": 1.1228,
|
| 98 |
+
"eval_samples_per_second": 4.453,
|
| 99 |
+
"eval_steps_per_second": 1.781,
|
| 100 |
+
"step": 24
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 7.142857142857143,
|
| 104 |
+
"grad_norm": 0.3781011248380884,
|
| 105 |
+
"learning_rate": 0.0001,
|
| 106 |
+
"loss": 0.1683,
|
| 107 |
+
"step": 25
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 8.0,
|
| 111 |
+
"eval_loss": 0.8222058415412903,
|
| 112 |
+
"eval_runtime": 1.1218,
|
| 113 |
+
"eval_samples_per_second": 4.457,
|
| 114 |
+
"eval_steps_per_second": 1.783,
|
| 115 |
+
"step": 28
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 8.571428571428571,
|
| 119 |
+
"grad_norm": 0.264128562506645,
|
| 120 |
+
"learning_rate": 6.173165676349103e-05,
|
| 121 |
+
"loss": 0.0699,
|
| 122 |
+
"step": 30
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.857142857142858,
|
| 126 |
+
"eval_loss": 0.8684995770454407,
|
| 127 |
+
"eval_runtime": 1.1229,
|
| 128 |
+
"eval_samples_per_second": 4.453,
|
| 129 |
+
"eval_steps_per_second": 1.781,
|
| 130 |
+
"step": 31
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 10.0,
|
| 134 |
+
"grad_norm": 0.24436285012704104,
|
| 135 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 136 |
+
"loss": 0.0369,
|
| 137 |
+
"step": 35
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 10.0,
|
| 141 |
+
"eval_loss": 0.906228244304657,
|
| 142 |
+
"eval_runtime": 1.122,
|
| 143 |
+
"eval_samples_per_second": 4.456,
|
| 144 |
+
"eval_steps_per_second": 1.782,
|
| 145 |
+
"step": 35
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 10.857142857142858,
|
| 149 |
+
"eval_loss": 0.9267812967300415,
|
| 150 |
+
"eval_runtime": 1.1231,
|
| 151 |
+
"eval_samples_per_second": 4.452,
|
| 152 |
+
"eval_steps_per_second": 1.781,
|
| 153 |
+
"step": 38
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 11.428571428571429,
|
| 157 |
+
"grad_norm": 0.13496685015110904,
|
| 158 |
+
"learning_rate": 7.612046748871327e-06,
|
| 159 |
+
"loss": 0.0248,
|
| 160 |
+
"step": 40
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 12.0,
|
| 164 |
+
"eval_loss": 0.9376745223999023,
|
| 165 |
+
"eval_runtime": 1.1206,
|
| 166 |
+
"eval_samples_per_second": 4.462,
|
| 167 |
+
"eval_steps_per_second": 1.785,
|
| 168 |
+
"step": 42
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 12.857142857142858,
|
| 172 |
+
"grad_norm": 0.13014521278503757,
|
| 173 |
+
"learning_rate": 0.0,
|
| 174 |
+
"loss": 0.0212,
|
| 175 |
+
"step": 45
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 12.857142857142858,
|
| 179 |
+
"eval_loss": 0.939548134803772,
|
| 180 |
+
"eval_runtime": 1.1886,
|
| 181 |
+
"eval_samples_per_second": 4.207,
|
| 182 |
+
"eval_steps_per_second": 1.683,
|
| 183 |
+
"step": 45
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 12.857142857142858,
|
| 187 |
+
"step": 45,
|
| 188 |
+
"total_flos": 10453480636416.0,
|
| 189 |
+
"train_loss": 0.2976339568694433,
|
| 190 |
+
"train_runtime": 339.6314,
|
| 191 |
+
"train_samples_per_second": 1.148,
|
| 192 |
+
"train_steps_per_second": 0.132
|
| 193 |
+
}
|
| 194 |
+
],
|
| 195 |
+
"logging_steps": 5,
|
| 196 |
+
"max_steps": 45,
|
| 197 |
+
"num_input_tokens_seen": 0,
|
| 198 |
+
"num_train_epochs": 15,
|
| 199 |
+
"save_steps": 1000,
|
| 200 |
+
"stateful_callbacks": {
|
| 201 |
+
"TrainerControl": {
|
| 202 |
+
"args": {
|
| 203 |
+
"should_epoch_stop": false,
|
| 204 |
+
"should_evaluate": false,
|
| 205 |
+
"should_log": false,
|
| 206 |
+
"should_save": true,
|
| 207 |
+
"should_training_stop": true
|
| 208 |
+
},
|
| 209 |
+
"attributes": {}
|
| 210 |
+
}
|
| 211 |
+
},
|
| 212 |
+
"total_flos": 10453480636416.0,
|
| 213 |
+
"train_batch_size": 4,
|
| 214 |
+
"trial_name": null,
|
| 215 |
+
"trial_params": null
|
| 216 |
+
}
|