tlocvsdyspneaTask-unsup-Qwen3-1.7B-datav3-all

This model is a fine-tuned version of ferrazzipietro/unsup-Qwen3-1.7B-datav3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3902
• F1 Micro: 0.6537
• F1 Macro: 0.2271
• F1 Weighted: 0.7372

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: 0.0003
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-07 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	F1 Macro	F1 Weighted
1.9287	0.1053	2	1.9820	0.0	0.0	0.0
1.9287	0.2105	4	1.9755	0.0	0.0	0.0
1.9588	0.3158	6	1.9457	0.0	0.0	0.0
1.9588	0.4211	8	1.8763	0.0	0.0	0.0
1.8472	0.5263	10	1.8017	0.0	0.0	0.0
1.8472	0.6316	12	1.7313	0.0	0.0	0.0
1.8472	0.7368	14	1.6685	0.0	0.0	0.0
1.7235	0.8421	16	1.6153	0.0	0.0	0.0
1.7235	0.9474	18	1.5706	0.0	0.0	0.0
1.5909	1.0526	20	1.5334	0.0	0.0	0.0
1.5909	1.1579	22	1.5028	0.0	0.0	0.0
1.5909	1.2632	24	1.4801	0.0	0.0	0.0
1.486	1.3684	26	1.4618	0.0019	0.0009	0.0039
1.486	1.4737	28	1.4464	0.1459	0.0449	0.2477
1.4147	1.5789	30	1.4351	0.4300	0.1275	0.4955
1.4147	1.6842	32	1.4249	0.4436	0.1711	0.5291
1.4147	1.7895	34	1.4166	0.6362	0.2293	0.7367
1.3675	1.8947	36	1.4107	0.6284	0.1914	0.6593
1.3675	2.0	38	1.4056	0.6362	0.2087	0.6968
1.3983	2.1053	40	1.4014	0.6148	0.2407	0.7488
1.3983	2.2105	42	1.3983	0.5058	0.2283	0.6510
1.3983	2.3158	44	1.3958	0.5019	0.2265	0.6402
1.3487	2.4211	46	1.3936	0.5428	0.2377	0.7010
1.3487	2.5263	48	1.3921	0.6284	0.2413	0.7547
1.4027	2.6316	50	1.3911	0.6498	0.2328	0.7469
1.4027	2.7368	52	1.3905	0.6556	0.2288	0.7411
1.4027	2.8421	54	1.3902	0.6459	0.2244	0.7301
1.2976	2.9474	56	1.3902	0.6537	0.2271	0.7372

Framework versions

• PEFT 0.18.1
• Transformers 4.51.0
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.0