See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: katuni4ka/tiny-random-qwen1.5-moe
bf16: auto
dataset_prepared_path: null
datasets:
- data_files:
  - bab283e818c97851_train_data.json
  ds_type: json
  format: custom
  path: bab283e818c97851_train_data.json
  type:
    field: null
    field_input: null
    field_instruction: context
    field_output: level_2
    field_system: null
    format: null
    no_input_format: null
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
evals_per_epoch: 3
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
group_by_length: false
hub_model_id: taopanda/2c157ea5-df82-4f8e-a01c-eed7d3c6cb1d
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
micro_batch_size: 1
num_epochs: 1
optimizer: paged_adamw_8bit
output_dir: ./outputs/out/taopanda-2_a4cfd74d-319f-41fe-90ca-a0914a9d703e
pad_to_sequence_len: false
resume_from_checkpoint: null
sample_packing: false
saves_per_epoch: 1
seed: 11975
sequence_len: 1024
special_tokens: null
strict: false
tf32: true
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: fatcat87-taopanda
wandb_log_model: null
wandb_mode: online
wandb_name: taopanda-2_a4cfd74d-319f-41fe-90ca-a0914a9d703e
wandb_project: subnet56
wandb_runid: taopanda-2_a4cfd74d-319f-41fe-90ca-a0914a9d703e
wandb_watch: null
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

2c157ea5-df82-4f8e-a01c-eed7d3c6cb1d

This model is a fine-tuned version of katuni4ka/tiny-random-qwen1.5-moe on the None dataset. It achieves the following results on the evaluation set:

• Loss: 11.7507

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: 1
• eval_batch_size: 1
• seed: 11975
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• total_eval_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
11.9292	0.0002	1	11.9319
11.7579	0.3334	1716	11.7547
11.7407	0.6668	3432	11.7507

Framework versions

• PEFT 0.11.1
• Transformers 4.42.3
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1