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:
  - 39d8d66c677d78b0_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/39d8d66c677d78b0_train_data.json
  type:
    field_input: Company Name
    field_instruction: Position
    field_output: Long Description
    format: '{instruction} {input}'
    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/42ebee07-be7a-4ee5-a174-a2e952f904a3
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: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 3600
micro_batch_size: 4
mlflow_experiment_name: /tmp/39d8d66c677d78b0_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: 0.035436614527594494
wandb_entity: null
wandb_mode: online
wandb_name: 7533f455-a96b-4817-83ad-f039208a2642
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 7533f455-a96b-4817-83ad-f039208a2642
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

42ebee07-be7a-4ee5-a174-a2e952f904a3

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: 1.7684

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: 3600

Training results

Training Loss	Epoch	Step	Validation Loss
2.2992	0.0002	1	2.3467
2.0978	0.0235	100	2.0582
1.9372	0.0470	200	2.0125
2.0753	0.0705	300	1.9854
1.9778	0.0940	400	1.9623
2.0292	0.1176	500	1.9422
2.0648	0.1411	600	1.9280
1.939	0.1646	700	1.9142
1.9822	0.1881	800	1.9019
1.9189	0.2116	900	1.8910
1.8554	0.2351	1000	1.8794
1.865	0.2586	1100	1.8685
1.9085	0.2821	1200	1.8599
1.8868	0.3057	1300	1.8505
1.9702	0.3292	1400	1.8422
1.7692	0.3527	1500	1.8342
1.8789	0.3762	1600	1.8269
1.8834	0.3997	1700	1.8199
1.8589	0.4232	1800	1.8144
1.9127	0.4467	1900	1.8083
1.7813	0.4702	2000	1.8032
1.7653	0.4938	2100	1.7981
1.8232	0.5173	2200	1.7945
1.8285	0.5408	2300	1.7902
1.8234	0.5643	2400	1.7860
1.7402	0.5878	2500	1.7825
1.7448	0.6113	2600	1.7795
1.8133	0.6348	2700	1.7770
1.7096	0.6583	2800	1.7748
1.7819	0.6819	2900	1.7728
1.7897	0.7054	3000	1.7714
1.7928	0.7289	3100	1.7701
1.8169	0.7524	3200	1.7693
1.8176	0.7759	3300	1.7689
1.8205	0.7994	3400	1.7685
1.7652	0.8229	3500	1.7684
1.8443	0.8464	3600	1.7684

Framework versions

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