See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: NousResearch/Yarn-Llama-2-13b-64k
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 113edc9ac50140a9_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/113edc9ac50140a9_train_data.json
  type:
    field_input: context
    field_instruction: question-X
    field_output: answer-Y
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: Romain-XV/2acb31f5-764e-45b9-8e74-92d66daffe15
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.00025
load_best_model_at_end: true
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 64
lora_dropout: 0.1
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
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 840
micro_batch_size: 4
mlflow_experiment_name: /tmp/113edc9ac50140a9_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
save_steps: 100
sequence_len: 2048
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.04
wandb_entity: null
wandb_mode: online
wandb_name: ff4eacc6-39f5-47c3-942d-394eedd249f4
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: ff4eacc6-39f5-47c3-942d-394eedd249f4
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

2acb31f5-764e-45b9-8e74-92d66daffe15

This model is a fine-tuned version of NousResearch/Yarn-Llama-2-13b-64k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4602

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: 0.00025
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• 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: 840

Training results

Training Loss	Epoch	Step	Validation Loss
27.5358	0.0010	1	3.4175
12.1336	0.0996	100	1.5528
12.1681	0.1992	200	1.5459
13.1276	0.2988	300	1.5165
10.6359	0.3984	400	1.5056
11.4921	0.4979	500	1.4838
11.1868	0.5975	600	1.4700
11.2614	0.6971	700	1.4632
11.4508	0.7967	800	1.4602

Framework versions

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