See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 9968ceef57d41e71_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/9968ceef57d41e71_train_data.json
  type:
    field_instruction: prompt
    field_output: gold_standard_solution
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
device_map:
  ? ''
  : 0,1,2,3,4,5,6,7
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: true
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: Alphatao/33d93d73-bbeb-404d-84dd-ffdded5566fb
hub_repo: null
hub_strategy: null
hub_token: null
learning_rate: 0.0002
load_best_model_at_end: true
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 128
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 2520
micro_batch_size: 4
mlflow_experiment_name: /tmp/9968ceef57d41e71_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
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: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 3ea4cbe7-e175-41c3-974d-7a7a8d18b3b7
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 3ea4cbe7-e175-41c3-974d-7a7a8d18b3b7
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

33d93d73-bbeb-404d-84dd-ffdded5566fb

This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-Chat-v1.0 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2245

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: 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: 2520

Training results

Training Loss	Epoch	Step	Validation Loss
0.3265	0.0005	1	0.3180
0.1697	0.0490	100	0.2371
0.236	0.0980	200	0.2349
0.2743	0.1470	300	0.2344
0.161	0.1960	400	0.2344
0.2298	0.2450	500	0.2332
0.3038	0.2940	600	0.2325
0.359	0.3430	700	0.2315
0.242	0.3920	800	0.2307
0.2286	0.4410	900	0.2303
0.2507	0.4900	1000	0.2292
0.2713	0.5390	1100	0.2284
0.2441	0.5880	1200	0.2277
0.2262	0.6370	1300	0.2266
0.1666	0.6860	1400	0.2256
0.3028	0.7350	1500	0.2250
0.1431	0.7840	1600	0.2243
0.2367	0.8330	1700	0.2237
0.2339	0.8820	1800	0.2230
0.1972	0.9310	1900	0.2227
0.2668	0.9800	2000	0.2221
0.2181	1.0290	2100	0.2238
0.1574	1.0780	2200	0.2245

Framework versions

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