See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: NousResearch/Yarn-Llama-2-7b-128k
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 4f01bd2925bc8a13_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/4f01bd2925bc8a13_train_data.json
  type:
    field_instruction: Title
    field_output: Genre
    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: 5
flash_attention: false
fp16: true
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: fedovtt/08c45d92-b447-4bdb-87ed-81e5e7053752
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: 32
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 74GiB
max_steps: 50
micro_batch_size: 4
mlflow_experiment_name: /tmp/4f01bd2925bc8a13_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 25
save_strategy: steps
sequence_len: 4056
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 08c45d92-b447-4bdb-87ed-81e5e7053752
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 08c45d92-b447-4bdb-87ed-81e5e7053752
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

08c45d92-b447-4bdb-87ed-81e5e7053752

This model is a fine-tuned version of NousResearch/Yarn-Llama-2-7b-128k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3162

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_TORCH 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss
13.5674	0.0343	1	1.6728
11.7391	0.1373	4	1.4845
7.9807	0.2747	8	0.9227
6.5926	0.4120	12	0.6636
3.9636	0.5494	16	0.5392
3.0476	0.6867	20	0.4746
3.0432	0.8240	24	0.4330
3.4371	0.9614	28	0.4020
2.1396	1.0987	32	0.3588
1.6073	1.2361	36	0.3393
3.2822	1.3734	40	0.3231
1.7314	1.5107	44	0.3193
1.7266	1.6481	48	0.3162

Framework versions

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