Modelpublic

rayonlabs/Qwen2_5-3B-Instruct-medical-o1-reasoning-SFT-9d95b060-41cc-4d8a-b03d-190792edce50

sourceHugging Faceotherupdated 2y agoView on Hugging Face

0likes6downloads

Model Card

<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/> <details><summary>See axolotl config</summary>

axolotl version: 0.4.1

yaml

adapter: lora
base_model: Qwen/Qwen2.5-3B-Instruct
bf16: auto
dataset_prepared_path: null
datasets:
- data_files:
  - a9370d8bc6e0e8be_train_data.json
  ds_type: json
  format: custom
  path: /root/G.O.D-test/core/data/a9370d8bc6e0e8be_train_data.json
  type:
    field_input: Complex_CoT
    field_instruction: Question
    field_output: Response
    format: '{instruction} {input}'
    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: 10
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: false
hub_model_id: souging/702609de-5f88-479b-8c42-393d171935f2
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: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_steps: 500
micro_batch_size: 4
mlflow_experiment_name: /tmp/a9370d8bc6e0e8be_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 4
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 10
sequence_len: 1024
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: 9d95b060-41cc-4d8a-b03d-190792edce50
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 9d95b060-41cc-4d8a-b03d-190792edce50
warmup_steps: 100
weight_decay: 0.01
xformers_attention: null

</details><br>

702609de-5f88-479b-8c42-393d171935f2

This model is a fine-tuned version of Qwen/Qwen2.5-3B-Instruct on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.7429

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
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—distributed_type: multi-GPU
—num_devices: 8
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 128
—totalevalbatch_size: 32
—optimizer: Use OptimizerNames.ADAMWBNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: cosine
—lrschedulerwarmup_steps: 100
—training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
1.0533	0.0055	1	1.2178
1.1372	0.0712	13	1.1764
0.9056	0.1425	26	0.8935
0.7864	0.2137	39	0.8048
0.796	0.2849	52	0.7839
0.7511	0.3562	65	0.7741
0.7576	0.4274	78	0.7689
0.7403	0.4986	91	0.7618
0.7916	0.5699	104	0.7577
0.7765	0.6411	117	0.7547
0.7039	0.7123	130	0.7529
0.7594	0.7836	143	0.7501
0.7966	0.8548	156	0.7477
0.8105	0.9260	169	0.7468
0.6799	0.9973	182	0.7456
0.7117	1.0685	195	0.7446
0.7082	1.1397	208	0.7479
0.7889	1.2110	221	0.7464
0.6094	1.2822	234	0.7451
0.7105	1.3534	247	0.7445
0.7257	1.4247	260	0.7432
0.7152	1.4959	273	0.7416
0.7657	1.5671	286	0.7402
0.7808	1.6384	299	0.7396
0.7059	1.7096	312	0.7394
0.7537	1.7808	325	0.7377
0.6318	1.8521	338	0.7373
0.7712	1.9233	351	0.7369
0.7703	1.9945	364	0.7362
0.6556	2.0658	377	0.7415
0.6248	2.1370	390	0.7405
0.6789	2.2082	403	0.7426
0.6595	2.2795	416	0.7436
0.6761	2.3507	429	0.7431
0.6399	2.4219	442	0.7434
0.6672	2.4932	455	0.7434
0.5704	2.5644	468	0.7433
0.6552	2.6356	481	0.7429
0.6455	2.7068	494	0.7429

Framework versions

—PEFT 0.13.2
—Transformers 4.46.0
—Pytorch 2.5.0+cu124
—Datasets 3.0.1
—Tokenizers 0.20.3