Hugging Face's logo

vdaita
/
diff-deepseek-line-changes

PEFT
mistral
Generated from Trainer
8-bit precision
bitsandbytes
Model card Files Community
diff-deepseek-line-changes / README.md
vdaita's picture
vdaita
Upload folder using huggingface_hub
e1c0be6 verified
preview code
|
raw
history blame
3.58 kB
metadata
license: apache-2.0
library_name: peft
tags:
  - generated_from_trainer
base_model: mistralai/Mistral-7B-v0.3
model-index:
  - name: outputs/axolotl-qlora-out-line
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.0

base_model: mistralai/Mistral-7B-v0.3
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer

load_in_8bit: true
load_in_4bit: false
strict: false

datasets:
  - path: vdaita/editpackft_inst_line
    type: oasst
dataset_prepared_path: 
val_set_size: 0.05
output_dir: ./outputs/axolotl-qlora-out-line

adapter: lora
lora_model_dir: 

sequence_len: 2048
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: 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: huggingface
wandb_log_model: axolotl-qlora-line-mistral

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 4
optimizer: paged_adamw_32bit
lr_scheduler: cosine
learning_rate: 0.0002

train_on_inputs: false
group_by_length: false
bf16: auto
fp16: 
tf32: false

gradient_checkpointing: true
logging_steps: 1
flash_attention: true

warmup_steps: 10
evals_per_epoch: 4
saves_per_epoch: 1

outputs/axolotl-qlora-out-line

This model is a fine-tuned version of mistralai/Mistral-7B-v0.3 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2883

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
  • distributed_type: multi-GPU
  • num_devices: 2
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • total_eval_batch_size: 4
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 10
  • num_epochs: 4

Training results

Training Loss Epoch Step Validation Loss
0.847 0.01 1 0.9975
0.3533 0.26 20 0.2772
0.299 0.52 40 0.2501
0.2288 0.77 60 0.2439
0.334 1.01 80 0.2394
0.3017 1.27 100 0.2399
0.2394 1.53 120 0.2416
0.2261 1.78 140 0.2400
0.177 2.02 160 0.2388
0.1911 2.28 180 0.2557
0.1884 2.54 200 0.2601
0.1516 2.79 220 0.2627
0.1545 3.03 240 0.2628
0.092 3.29 260 0.2915
0.1251 3.55 280 0.2892
0.109 3.8 300 0.2883

Framework versions

  • PEFT 0.10.0
  • Transformers 4.40.0.dev0
  • Pytorch 2.3.0+cu121
  • Datasets 2.15.0
  • Tokenizers 0.15.0