See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: peft-internal-testing/tiny-dummy-qwen2
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 99a1d9d467a445dc_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/99a1d9d467a445dc_train_data.json
  type:
    field_instruction: question
    field_output: answer
    format: '{instruction}'
    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: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
gradient_clipping: 1.0
group_by_length: false
hub_model_id: oldiday/8d08a67f-fb94-4323-bc8e-2d99913d5901
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: 0
logging_steps: 3
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_steps: 100
micro_batch_size: 8
mlflow_experiment_name: /tmp/99a1d9d467a445dc_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
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: techspear-hub
wandb_mode: online
wandb_name: 2f1a76ed-541f-4b8a-8491-311724d62463
wandb_project: Gradients-On-Six
wandb_run: your_name
wandb_runid: 2f1a76ed-541f-4b8a-8491-311724d62463
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

8d08a67f-fb94-4323-bc8e-2d99913d5901

This model is a fine-tuned version of peft-internal-testing/tiny-dummy-qwen2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 11.9119

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: 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: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0060	1	11.9395
11.9385	0.0541	9	11.9390
11.9374	0.1081	18	11.9375
11.9358	0.1622	27	11.9353
11.9325	0.2162	36	11.9319
11.9281	0.2703	45	11.9272
11.9226	0.3243	54	11.9219
11.9178	0.3784	63	11.9173
11.915	0.4324	72	11.9143
11.9134	0.4865	81	11.9127
11.9126	0.5405	90	11.9120
11.9124	0.5946	99	11.9119

Framework versions

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