See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: The-matt/llama2_ko-7b_distinctive-snowflake-182_1060
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - d62dc4319fb018dc_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/d62dc4319fb018dc_train_data.json
  type:
    field_input: wnli_sent2
    field_instruction: wnli_sent1
    field_output: sentence
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 4
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
gradient_clipping: 1.0
group_by_length: false
hub_model_id: brixeus/211afcf5-69d4-43a9-9ad9-fd52b1005add
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: false
local_rank: 0
logging_steps: 3
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_steps: 100
micro_batch_size: 8
mlflow_experiment_name: /tmp/d62dc4319fb018dc_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 4
sequence_len: 1024
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: techspear-hub
wandb_mode: online
wandb_name: 7ec91b99-cf37-41fd-9d4e-15c9b3abe2f0
wandb_project: Gradients-On-Three
wandb_run: your_name
wandb_runid: 7ec91b99-cf37-41fd-9d4e-15c9b3abe2f0
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

211afcf5-69d4-43a9-9ad9-fd52b1005add

This model is a fine-tuned version of The-matt/llama2_ko-7b_distinctive-snowflake-182_1060 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0072

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• 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: 10
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0126	1	0.4468
0.3273	0.1136	9	0.1988
0.0622	0.2271	18	0.0430
0.0257	0.3407	27	0.0213
0.0079	0.4543	36	0.0156
0.0072	0.5678	45	0.0128
0.0063	0.6814	54	0.0099
0.0067	0.7950	63	0.0085
0.0062	0.9085	72	0.0077
0.0064	1.0221	81	0.0074
0.0054	1.1356	90	0.0072
0.0047	1.2492	99	0.0072

Framework versions

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