lemexp-task1-template-small-r16-18ep-8bit

yalhessi/lemexp-task1-template-small-r16-18ep-8bit

lemexp-task1-template-small-r16-18ep-8bit is a machine learning model from yalhessi. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft. The card lists the license as other.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

20% of all-time downloads

All-time downloads

Public

Repo size

585 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors585 MB · 100%

At a glance

Library: peft
License: other
Access: Public
Created: Jul 10, 2026
Updated: Jul 10, 2026
SHA: 75e939bf

Base models

deepseek-ai/deepseek-coder-1.3b-base

Model card

From the Hugging Face model README

lemexp-task1-template-small-r16-18ep-8bit

This model is a fine-tuned version of deepseek-ai/deepseek-coder-1.3b-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1938

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: 0.0008
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 32
total_eval_batch_size: 32
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: linear
lr_scheduler_warmup_ratio: 0.03
num_epochs: 18
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.2	360	0.3005
0.4769	0.4	720	0.2573
0.2634	0.6	1080	0.2495
0.2634	0.8	1440	0.2463
0.2424	1.0	1800	0.2241
0.219	1.2	2160	0.2199
0.2087	1.4	2520	0.2042
0.2087	1.6	2880	0.2102
0.2079	1.8	3240	0.2040
0.1998	2.0	3600	0.2107
0.1998	2.2	3960	0.1998
0.185	2.4	4320	0.1915
0.1809	2.6	4680	0.1827
0.1769	2.8	5040	0.1887
0.1769	3.0	5400	0.1943
0.1779	3.2	5760	0.1844
0.1661	3.4	6120	0.1804
0.1661	3.6	6480	0.1760
0.1638	3.8	6840	0.1747
0.1634	4.0	7200	0.1766
0.1529	4.2	7560	0.1767
0.1529	4.4	7920	0.1713
0.1546	4.6	8280	0.1720
0.1514	4.8	8640	0.1690
0.1495	5.0	9000	0.1717
0.1495	5.2	9360	0.1656
0.1375	5.4	9720	0.1661
0.1409	5.6	10080	0.1686
0.1409	5.8	10440	0.1584
0.142	6.0	10800	0.1550
0.1324	6.2	11160	0.1595
0.1274	6.4	11520	0.1581
0.1274	6.6	11880	0.1542
0.1304	6.8	12240	0.1493
0.1279	7.0	12600	0.1496
0.1279	7.2	12960	0.1552
0.1166	7.4	13320	0.1515
0.1177	7.6	13680	0.1566
0.1184	7.8	14040	0.1455
0.1184	8.0	14400	0.1472
0.1159	8.2	14760	0.1480
0.106	8.4	15120	0.1476
0.106	8.6	15480	0.1482
0.1097	8.8	15840	0.1482
0.1084	9.0	16200	0.1413
0.0991	9.2	16560	0.1478
0.0991	9.4	16920	0.1517
0.0982	9.6	17280	0.1430
0.0998	9.8	17640	0.1406
0.0988	10.0	18000	0.1389
0.0988	10.2	18360	0.1487
0.0844	10.4	18720	0.1445
0.0882	10.6	19080	0.1410
0.0882	10.8	19440	0.1428
0.0888	11.0	19800	0.1408
0.0842	11.2	20160	0.1477
0.0784	11.4	20520	0.1434
0.0784	11.6	20880	0.1471
0.0806	11.8	21240	0.1433
0.0804	12.0	21600	0.1396
0.0804	12.2	21960	0.1497
0.0701	12.4	22320	0.1456
0.0713	12.6	22680	0.1461
0.0721	12.8	23040	0.1452
0.0721	13.0	23400	0.1434
0.0702	13.2	23760	0.1569
0.0632	13.4	24120	0.1545
0.0632	13.6	24480	0.1487
0.0647	13.8	24840	0.1483
0.0641	14.0	25200	0.1443
0.0588	14.2	25560	0.1547
0.0588	14.4	25920	0.1559
0.0569	14.6	26280	0.1521
0.0579	14.8	26640	0.1527
0.0583	15.0	27000	0.1533
0.0583	15.2	27360	0.1706
0.0518	15.4	27720	0.1639
0.0517	15.6	28080	0.1641
0.0517	15.8	28440	0.1627
0.0526	16.0	28800	0.1652
0.0505	16.2	29160	0.1785
0.0478	16.4	29520	0.1765
0.0478	16.6	29880	0.1783
0.0486	16.8	30240	0.1764
0.0479	17.0	30600	0.1750
0.0479	17.2	30960	0.1925
0.0454	17.4	31320	0.1948
0.0451	17.6	31680	0.1935
0.0452	17.8	32040	0.1951
0.0452	18.0	32400	0.1938

Framework versions

PEFT 0.14.0
Transformers 4.47.0
Pytorch 2.7.1+cu128
Datasets 4.2.0
Tokenizers 0.21.4

Library: peft
License: other
Created: Jul 10, 2026
Updated: Jul 10, 2026