adapter: lora
base_model: teknium/OpenHermes-2.5-Mistral-7B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 9edc46c970298de0_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/
  type:
    field_instruction: instruct
    field_output: output
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
device_map: auto
dynamic_lora_per_layer: true
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
evals_per_epoch: null
evaluation_strategy: steps
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: true
group_by_length: false
hub_model_id: JoshMe1/e699156c-4f55-4d48-960e-1b1d0dec0492
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: true
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 128
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_finder: true
lr_scheduler: cosine
lr_scheduler_args: []
max_memory:
  0: 130GB
max_steps: 1500
micro_batch_size: 8
mixed_precision: bf16
mlflow_experiment_name: /tmp/9edc46c970298de0_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 4
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
save_strategy: steps
save_total_limit: 3
saves_per_epoch: null
scheduler:
  factor: 0.5
  monitor: eval_loss
  patience: 1
  threshold: 0.005
  type: ReduceLROnPlateau
sequence_len: 2048
special_tokens:
  pad_token: <|im_end|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
training_stages:
- learning_rate: 0.0002
  name: warmup
  num_train_epochs: 1
- learning_rate: 2.0e-05
  name: main
trl:
  ema: true
  ema_decay: 0.999
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 280c19eb-1ad3-4a75-8b48-0179144a1d93
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 280c19eb-1ad3-4a75-8b48-0179144a1d93
warmup_steps: 150
weight_decay: 0.01
xformers_attention: true

e699156c-4f55-4d48-960e-1b1d0dec0492

This model is a fine-tuned version of teknium/OpenHermes-2.5-Mistral-7B on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.7948

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: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
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: 150
training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0001	1	0.9281
1.3618	0.0057	100	0.7266
1.5356	0.0115	200	0.7850
1.5832	0.0172	300	0.7948

Framework versions

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