Hugging Face's logo

lesso17
/
4de6d45f-421b-462f-8f05-b63f7c8fdea1

PEFT
Safetensors
mixtral
axolotl
Generated from Trainer
Model card Files Community
4de6d45f-421b-462f-8f05-b63f7c8fdea1 / README.md
lesso17's picture
lesso17
End of training
99c6a55 verified
preview code
|
raw
history blame
4.08 kB
metadata
library_name: peft
base_model: TitanML/tiny-mixtral
tags:
  - axolotl
  - generated_from_trainer
model-index:
  - name: 4de6d45f-421b-462f-8f05-b63f7c8fdea1
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.1

adapter: lora
auto_find_batch_size: true
base_model: TitanML/tiny-mixtral
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 4322762c82fd69e0_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/4322762c82fd69e0_train_data.json
  type:
    field_instruction: prompt
    field_output: response
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
do_eval: true
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 50
evals_per_epoch: null
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: false
group_by_length: true
hub_model_id: lesso17/4de6d45f-421b-462f-8f05-b63f7c8fdea1
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.000217
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 128
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 500
micro_batch_size: 4
mlflow_experiment_name: /tmp/G.O.D/4322762c82fd69e0_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 50
saves_per_epoch: null
seed: 170
sequence_len: 512
special_tokens:
  pad_token: </s>
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 5333f4b9-6be8-4394-8c0f-19c9bfdc6d30
wandb_project: 17a
wandb_run: your_name
wandb_runid: 5333f4b9-6be8-4394-8c0f-19c9bfdc6d30
warmup_steps: 50
weight_decay: 0.0
xformers_attention: null

4de6d45f-421b-462f-8f05-b63f7c8fdea1

This model is a fine-tuned version of TitanML/tiny-mixtral on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 8.3056

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: 0.000217
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 170
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 8
  • optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 50
  • training_steps: 392

Training results

Training Loss Epoch Step Validation Loss
No log 0.0026 1 10.5671
8.9191 0.1277 50 8.9119
8.654 0.2554 100 8.7470
8.3712 0.3831 150 8.5024
8.3764 0.5109 200 8.4021
8.3127 0.6386 250 8.3437
8.2662 0.7663 300 8.3173
8.2846 0.8940 350 8.3056

Framework versions

  • PEFT 0.13.2
  • Transformers 4.46.0
  • Pytorch 2.5.0+cu124
  • Datasets 3.0.1
  • Tokenizers 0.20.1