You can deploy this model and start earning money today!

airesearch/
Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405

Name: Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405
Availability: InStock

generated_from_trainer

Qwen/Qwen3-32B

adapter

other

Views

No views yet

Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405

This model is a fine-tuned version of Qwen/Qwen3-32B on the alpaca-th-52k, the dolly-th-15k and the wangchan-instruct datasets. It achieves the following results on the evaluation set:

Loss: 0.6359

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: 2405
distributed_type: multi-GPU
num_devices: 32
gradient_accumulation_steps: 8
total_train_batch_size: 512
total_eval_batch_size: 64
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_ratio: 0.1
num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.9657	0.0575	10	1.0428
0.8056	0.1149	20	0.8197
0.7664	0.1724	30	0.7526
0.7675	0.2299	40	0.7330
0.7185	0.2874	50	0.7189
0.6613	0.3448	60	0.7070
0.7022	0.4023	70	0.6979
0.7143	0.4598	80	0.6901
0.6819	0.5172	90	0.6846
0.6613	0.5747	100	0.6784
0.6887	0.6322	110	0.6724
0.6812	0.6897	120	0.6693
0.677	0.7471	130	0.6659
0.6793	0.8046	140	0.6632
0.6701	0.8621	150	0.6610
0.6533	0.9195	160	0.6587
0.6666	0.9770	170	0.6571
0.6506	1.0345	180	0.6560
0.6385	1.0920	190	0.6541
0.6499	1.1494	200	0.6528
0.6515	1.2069	210	0.6511
0.6352	1.2644	220	0.6504
0.6337	1.3218	230	0.6487
0.6438	1.3793	240	0.6475
0.6529	1.4368	250	0.6465
0.6469	1.4943	260	0.6452
0.6087	1.5517	270	0.6444
0.6477	1.6092	280	0.6435
0.6302	1.6667	290	0.6425
0.6477	1.7241	300	0.6415
0.6245	1.7816	310	0.6405
0.6486	1.8391	320	0.6398
0.6376	1.8966	330	0.6389
0.6437	1.9540	340	0.6382
0.6161	2.0115	350	0.6380
0.6057	2.0690	360	0.6382
0.5957	2.1264	370	0.6382
0.6465	2.1839	380	0.6380
0.5811	2.2414	390	0.6378
0.6004	2.2989	400	0.6375
0.6484	2.3563	410	0.6370
0.5845	2.4138	420	0.6369
0.5928	2.4713	430	0.6365
0.5795	2.5287	440	0.6364
0.6306	2.5862	450	0.6360
0.6362	2.6437	460	0.6362
0.6316	2.7011	470	0.6361
0.6355	2.7586	480	0.6359
0.5977	2.8161	490	0.6359
0.6007	2.8736	500	0.6359
0.615	2.9310	510	0.6359
0.6153	2.9885	520	0.6359

Framework versions

PEFT 0.15.2
Transformers 4.52.3
Pytorch 2.7.0+cu126
Datasets 3.6.0
Tokenizers 0.21.1

airesearch/Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405

Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

airesearch/
Qwen3-32B-alpaca-th-52k-dolly-th-15k-wangchan-instruct-seed-2405