See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: NousResearch/Yarn-Llama-2-7b-64k
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - fd8cb244a58eab5a_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/fd8cb244a58eab5a_train_data.json
  type:
    field_input: input
    field_instruction: instruction
    field_output: output
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 256
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 32
gradient_checkpointing: true
group_by_length: false
hub_model_id: mamung/0bd73ea6-3942-4ea0-be21-e85825c1f22e
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 3
lora_alpha: 64
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
lr_scheduler: cosine
max_grad_norm: 2
max_steps: 100
micro_batch_size: 2
mlflow_experiment_name: /tmp/fd8cb244a58eab5a_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optim_args:
  adam_beta1: 0.9
  adam_beta2: 0.95
  adam_epsilon: 1.0e-05
optimizer: adamw_torch
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: 2048
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: eddysang
wandb_mode: online
wandb_name: 41c1cfc2-4854-41dd-aa2c-b18f0b6c6123
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 41c1cfc2-4854-41dd-aa2c-b18f0b6c6123
warmup_steps: 20
weight_decay: 0.02
xformers_attention: false

0bd73ea6-3942-4ea0-be21-e85825c1f22e

This model is a fine-tuned version of NousResearch/Yarn-Llama-2-7b-64k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6466

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: 32
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=adam_beta1=0.9,adam_beta2=0.95,adam_epsilon=1e-05
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 20
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0013	1	0.9565
27.3274	0.0117	9	0.8349
24.32	0.0234	18	0.7370
22.5239	0.0351	27	0.6974
21.4471	0.0468	36	0.6797
21.4632	0.0585	45	0.6686
21.577	0.0702	54	0.6620
21.3475	0.0819	63	0.6561
20.0823	0.0936	72	0.6524
20.8383	0.1053	81	0.6487
21.1298	0.1170	90	0.6471
20.3105	0.1287	99	0.6466

Framework versions

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