File size: 7,401 Bytes
3ae0e54 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 |
---
license: llama2
library_name: peft
tags:
- axolotl
- generated_from_trainer
base_model: codellama/CodeLlama-7b-hf
model-index:
- name: EvolCodeLlama-7b
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
<details><summary>See axolotl config</summary>
axolotl version: `0.4.0`
```yaml
base_model: codellama/CodeLlama-7b-hf
base_model_config: codellama/CodeLlama-7b-hf
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
is_llama_derived_model: true
hub_model_id: EvolCodeLlama-7b
load_in_8bit: false
load_in_4bit: true
strict: false
datasets:
- path: ptoro/Evol-Instruct-Python-1k-testing
type: alpaca
dataset_prepared_path: last_run_prepared
val_set_size: 0.02
output_dir: ./qlora-out
adapter: qlora
lora_model_dir:
sequence_len: 2048
sample_packing: true
lora_r: 32
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true
lora_fan_in_fan_out:
wandb_project: axolotl
wandb_entity:
wandb_watch:
wandb_run_id:
wandb_log_model:
gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 3
optimizer: paged_adamw_32bit
lr_scheduler: cosine
learning_rate: 0.0002
train_on_inputs: false
group_by_length: false
bf16: true
fp16: false
tf32: false
gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
warmup_steps: 100
eval_steps: 0.01
save_strategy: epoch
save_steps:
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
bos_token: "<s>"
eos_token: "</s>"
unk_token: "<unk>"
```
</details><br>
# EvolCodeLlama-7b
This model is a fine-tuned version of [codellama/CodeLlama-7b-hf](https://huggingface.co/codellama/CodeLlama-7b-hf) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3828
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.3627 | 0.01 | 1 | 0.5027 |
| 0.3412 | 0.03 | 4 | 0.5026 |
| 0.3806 | 0.07 | 8 | 0.5023 |
| 0.392 | 0.1 | 12 | 0.5018 |
| 0.4141 | 0.14 | 16 | 0.4999 |
| 0.3433 | 0.17 | 20 | 0.4954 |
| 0.3702 | 0.21 | 24 | 0.4851 |
| 0.2948 | 0.24 | 28 | 0.4682 |
| 0.3387 | 0.28 | 32 | 0.4499 |
| 0.2437 | 0.31 | 36 | 0.4331 |
| 0.2526 | 0.35 | 40 | 0.4221 |
| 0.2721 | 0.38 | 44 | 0.4146 |
| 0.2292 | 0.42 | 48 | 0.4089 |
| 0.1986 | 0.45 | 52 | 0.4028 |
| 0.3258 | 0.48 | 56 | 0.3983 |
| 0.3509 | 0.52 | 60 | 0.3950 |
| 0.2697 | 0.55 | 64 | 0.3926 |
| 0.2646 | 0.59 | 68 | 0.3907 |
| 0.3979 | 0.62 | 72 | 0.3900 |
| 0.2737 | 0.66 | 76 | 0.3880 |
| 0.2271 | 0.69 | 80 | 0.3865 |
| 0.247 | 0.73 | 84 | 0.3847 |
| 0.3112 | 0.76 | 88 | 0.3824 |
| 0.2724 | 0.8 | 92 | 0.3820 |
| 0.207 | 0.83 | 96 | 0.3814 |
| 0.3492 | 0.87 | 100 | 0.3810 |
| 0.2474 | 0.9 | 104 | 0.3802 |
| 0.4037 | 0.94 | 108 | 0.3785 |
| 0.2295 | 0.97 | 112 | 0.3773 |
| 0.2689 | 1.0 | 116 | 0.3760 |
| 0.2546 | 1.02 | 120 | 0.3753 |
| 0.1916 | 1.05 | 124 | 0.3768 |
| 0.2458 | 1.09 | 128 | 0.3758 |
| 0.2155 | 1.12 | 132 | 0.3768 |
| 0.2341 | 1.16 | 136 | 0.3773 |
| 0.1909 | 1.19 | 140 | 0.3793 |
| 0.1911 | 1.23 | 144 | 0.3759 |
| 0.2096 | 1.26 | 148 | 0.3761 |
| 0.2353 | 1.29 | 152 | 0.3772 |
| 0.2606 | 1.33 | 156 | 0.3773 |
| 0.1485 | 1.36 | 160 | 0.3778 |
| 0.1807 | 1.4 | 164 | 0.3749 |
| 0.2294 | 1.43 | 168 | 0.3770 |
| 0.216 | 1.47 | 172 | 0.3759 |
| 0.1791 | 1.5 | 176 | 0.3727 |
| 0.2605 | 1.54 | 180 | 0.3733 |
| 0.2838 | 1.57 | 184 | 0.3738 |
| 0.2632 | 1.61 | 188 | 0.3694 |
| 0.1839 | 1.64 | 192 | 0.3686 |
| 0.1939 | 1.68 | 196 | 0.3690 |
| 0.2413 | 1.71 | 200 | 0.3699 |
| 0.1494 | 1.74 | 204 | 0.3689 |
| 0.2782 | 1.78 | 208 | 0.3695 |
| 0.2314 | 1.81 | 212 | 0.3696 |
| 0.2499 | 1.85 | 216 | 0.3691 |
| 0.1976 | 1.88 | 220 | 0.3672 |
| 0.2587 | 1.92 | 224 | 0.3660 |
| 0.2598 | 1.95 | 228 | 0.3658 |
| 0.2686 | 1.99 | 232 | 0.3666 |
| 0.216 | 2.01 | 236 | 0.3673 |
| 0.1261 | 2.04 | 240 | 0.3723 |
| 0.1938 | 2.08 | 244 | 0.3811 |
| 0.1906 | 2.11 | 248 | 0.3869 |
| 0.1375 | 2.15 | 252 | 0.3829 |
| 0.228 | 2.18 | 256 | 0.3796 |
| 0.2524 | 2.22 | 260 | 0.3789 |
| 0.118 | 2.25 | 264 | 0.3809 |
| 0.2224 | 2.29 | 268 | 0.3834 |
| 0.1477 | 2.32 | 272 | 0.3847 |
| 0.2095 | 2.35 | 276 | 0.3849 |
| 0.1919 | 2.39 | 280 | 0.3820 |
| 0.1916 | 2.42 | 284 | 0.3804 |
| 0.1625 | 2.46 | 288 | 0.3788 |
| 0.2054 | 2.49 | 292 | 0.3794 |
| 0.1605 | 2.53 | 296 | 0.3810 |
| 0.1564 | 2.56 | 300 | 0.3819 |
| 0.196 | 2.6 | 304 | 0.3822 |
| 0.1975 | 2.63 | 308 | 0.3830 |
| 0.1406 | 2.67 | 312 | 0.3833 |
| 0.2754 | 2.7 | 316 | 0.3830 |
| 0.1544 | 2.74 | 320 | 0.3829 |
| 0.1733 | 2.77 | 324 | 0.3830 |
| 0.1862 | 2.81 | 328 | 0.3832 |
| 0.1634 | 2.84 | 332 | 0.3829 |
| 0.1966 | 2.87 | 336 | 0.3830 |
| 0.1306 | 2.91 | 340 | 0.3831 |
| 0.1444 | 2.94 | 344 | 0.3828 |
### Framework versions
- PEFT 0.7.2.dev0
- Transformers 4.37.0
- Pytorch 2.0.1+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0 |