See axolotl config

axolotl version: 0.6.0

adapter: lora
base_model: unsloth/Hermes-3-Llama-3.1-8B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 612dd4d7fe9a0b1e_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/612dd4d7fe9a0b1e_train_data.json
  type:
    field_input: document_title
    field_instruction: document_plaintext
    field_output: question_text
    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: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 16
gradient_checkpointing: false
group_by_length: true
hub_model_id: jssky/2b75c7b2-823e-4f5e-bf4f-8dcb81b1c8e5
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: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 1500
micro_batch_size: 2
mlflow_experiment_name: /tmp/612dd4d7fe9a0b1e_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
saves_per_epoch: 4
sequence_len: 1024
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: 7c4a3015-c0b6-4621-95f7-cc221790d18b
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 7c4a3015-c0b6-4621-95f7-cc221790d18b
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

2b75c7b2-823e-4f5e-bf4f-8dcb81b1c8e5

This model is a fine-tuned version of unsloth/Hermes-3-Llama-3.1-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6549

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: 16
• total_train_batch_size: 32
• optimizer: Use adamw_bnb_8bit 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: 1500

Training results

Training Loss	Epoch	Step	Validation Loss
0.5968	0.0978	375	0.7516
0.9123	0.1955	750	0.6993
0.7483	0.2933	1125	0.6648
0.9141	0.3911	1500	0.6549

Framework versions

• PEFT 0.14.0
• Transformers 4.46.3
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3