minpeter/LoRA-Qwen3-4b-v1-iteration-02-sf-apigen-01

See axolotl config

axolotl version: 0.9.2

base_model: Qwen/Qwen3-4B
hub_model_id: minpeter/LoRA-Qwen3-4b-v1-iteration-01-sf-apigen-01

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: minpeter/apigen-mt-5k-friendli
    data_files:
      - train.jsonl
      - test.jsonl
    type: chat_template
    roles_to_train: ["assistant"]
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    shards: 3
chat_template: chatml

dataset_prepared_path: last_run_prepared

output_dir: ./output

adapter: lora
lora_model_dir:

sequence_len: 8192
pad_to_sequence_len: true
sample_packing: true

val_set_size: 0.05
eval_sample_packing: true
evals_per_epoch: 3

lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj

wandb_project: "axolotl"
wandb_entity: "kasfiekfs-e"
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 2
micro_batch_size: 2
num_epochs: 2
optimizer: adamw_8bit
lr_scheduler: cosine
learning_rate: 0.0002

train_on_inputs: false
group_by_length: false
bf16: auto
tf32: true

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

loss_watchdog_threshold: 5.0
loss_watchdog_patience: 3

warmup_steps: 10
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:

LoRA-Qwen3-4b-v1-iteration-01-sf-apigen-01

This model is a fine-tuned version of Qwen/Qwen3-4B on the minpeter/apigen-mt-5k-friendli dataset. It achieves the following results on the evaluation set:

• Loss: 0.2285

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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_8BIT 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
• num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss
1.4432	0.0069	1	1.0528
0.3253	0.3322	48	0.2922
0.4198	0.6644	96	0.2638
0.4426	0.9965	144	0.2449
0.2287	1.3253	192	0.2340
0.1526	1.6574	240	0.2299
0.268	1.9896	288	0.2285

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1