See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: Qwen/Qwen2.5-1.5B
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 26e8230457ed8fb8_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/26e8230457ed8fb8_train_data.json
  type:
    field_input: endings
    field_instruction: ctx_a
    field_output: label
    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: false
fp16: true
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: ardaspear/db44bfcc-673a-4766-889d-92290f4d9cdf
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: 128
lora_dropout: 0.1
lora_fan_in_fan_out: true
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 72GB
max_steps: 200
micro_batch_size: 8
mlflow_experiment_name: /tmp/26e8230457ed8fb8_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: false
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: leixa-personal
wandb_mode: online
wandb_name: db44bfcc-673a-4766-889d-92290f4d9cdf
wandb_project: Gradients-On-Three
wandb_run: your_name
wandb_runid: db44bfcc-673a-4766-889d-92290f4d9cdf
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

db44bfcc-673a-4766-889d-92290f4d9cdf

This model is a fine-tuned version of Qwen/Qwen2.5-1.5B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5560

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: 200

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0006	1	6.7477
0.705	0.0096	17	0.7000
0.7011	0.0192	34	0.6982
0.6933	0.0287	51	0.6953
0.7026	0.0383	68	0.6970
0.6852	0.0479	85	0.6992
0.6876	0.0575	102	0.6745
0.6684	0.0671	119	0.6285
0.6905	0.0767	136	0.6011
0.6133	0.0862	153	0.5719
0.5274	0.0958	170	0.5606
0.5679	0.1054	187	0.5560

Framework versions

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