flan-t5la-small

hrezaei/flan-t5la-small

flan-t5la-small is a feature extraction model from hrezaei. Use it when you need embeddings to search or compare text. It is set up for transformers.

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

Downloads · 30 days

1% of all-time downloads

All-time downloads

1.4K

Public

Parameters

93.4M

40 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors374 MB · 99%

At a glance

Task: Feature Extraction
Library: transformers
Model type: t5la_adapter
Access: Public
Created: Oct 28, 2025
Updated: Jan 16, 2026
SHA: eeba7bcc

Datasets

HuggingFaceFW/fineweb

Model card

From the Hugging Face model README

flan-t5la-small

This model is a fine-tuned version of hrezaei/flan-t5la-small on the HuggingFaceFW/fineweb sample-350BT dataset. It achieves the following results on the evaluation set:

Perplexity: 7.7313
Loss: 2.0453
Accuracy: 0.0032
Lookahead Perplexity: 49.1518
Lookahead Loss: 3.8949
Base Perplexity: 1.2146
Base Loss: 0.1944

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: 5e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
distributed_type: multi-GPU
num_devices: 2
total_train_batch_size: 32
total_eval_batch_size: 32
optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
training_steps: 524288

Training results

Training Loss	Epoch	Step	Perplexity	Validation Loss	Accuracy	Lookahead Perplexity	Lookahead Loss	Base Perplexity	Base Loss
4.8052	0.0095	5000	250.1917	5.5222	0.0032	51687.8881	10.8530	1.2146	0.1944
4.1668	0.0191	10000	88.2828	4.4805	0.0032	6430.9941	8.7689	1.2146	0.1944
3.7641	0.0286	15000	48.8034	3.8878	0.0032	1966.0577	7.5838	1.2146	0.1944
3.5174	0.0381	20000	33.9964	3.5263	0.0032	953.9463	6.8606	1.2146	0.1944
3.3815	0.0477	25000	26.7476	3.2864	0.0032	590.2392	6.3805	1.2146	0.1944
3.2612	0.0572	30000	22.4741	3.1124	0.0032	416.4622	6.0318	1.2146	0.1944
3.1695	0.0668	35000	19.6661	2.9789	0.0032	318.7243	5.7643	1.2146	0.1944
3.1099	0.0763	40000	17.6862	2.8728	0.0032	257.6626	5.5517	1.2146	0.1944
3.0454	0.0858	45000	16.2331	2.7871	0.0032	216.9703	5.3798	1.2146	0.1944
2.9932	0.0954	50000	15.1058	2.7151	0.0032	187.8344	5.2356	1.2146	0.1944
2.9685	0.1049	55000	14.2347	2.6557	0.0032	166.7505	5.1165	1.2146	0.1944
2.9146	0.1144	60000	13.5266	2.6047	0.0032	150.5264	5.0141	1.2146	0.1944
2.8939	1.0048	65000	12.9238	2.5591	0.0032	137.3885	4.9228	1.2146	0.1944
2.8667	1.0143	70000	12.4200	2.5193	0.0032	126.8708	4.8432	1.2146	0.1944
2.8495	1.0238	75000	11.9951	2.4845	0.0032	118.3322	4.7735	1.2146	0.1944
2.8226	1.0334	80000	11.6344	2.4540	0.0032	111.3047	4.7123	1.2146	0.1944
2.8131	1.0429	85000	11.3094	2.4256	0.0032	105.1683	4.6556	1.2146	0.1944
2.7982	1.0525	90000	11.0316	2.4008	0.0032	100.0606	4.6058	1.2146	0.1944
2.7809	1.0620	95000	10.7938	2.3790	0.0032	95.7859	4.5621	1.2146	0.1944
2.7659	1.0715	100000	10.5758	2.3586	0.0032	91.9492	4.5212	1.2146	0.1944
2.7617	1.0811	105000	10.3792	2.3398	0.0032	88.5628	4.4837	1.2146	0.1944
2.7458	1.0906	110000	10.1936	2.3218	0.0032	85.4272	4.4477	1.2146	0.1944
2.7383	1.1001	115000	10.0342	2.3060	0.0032	82.7731	4.4161	1.2146	0.1944
2.7269	1.1097	120000	9.8958	2.2921	0.0032	80.4986	4.3882	1.2146	0.1944
2.7131	1.1192	125000	9.7566	2.2779	0.0032	78.2504	4.3599	1.2146	0.1944
2.6989	2.0095	130000	9.6413	2.2661	0.0032	76.4115	4.3361	1.2146	0.1944
2.7099	2.0191	135000	9.5230	2.2537	0.0032	74.5514	4.3115	1.2146	0.1944
2.7143	2.0286	140000	9.4245	2.2433	0.0032	73.0154	4.2907	1.2146	0.1944
2.682	2.0381	145000	9.3258	2.2328	0.0032	71.4968	4.2697	1.2146	0.1944
2.7051	2.0477	150000	9.2432	2.2239	0.0032	70.2333	4.2518	1.2146	0.1944
2.6935	2.0572	155000	9.1599	2.2148	0.0032	68.9760	4.2338	1.2146	0.1944
2.6733	2.0668	160000	9.0837	2.2065	0.0032	67.8339	4.2171	1.2146	0.1944
2.6706	2.0763	165000	9.0112	2.1985	0.0032	66.7535	4.2010	1.2146	0.1944
2.6538	2.0858	170000	8.9443	2.1910	0.0032	65.7671	4.1861	1.2146	0.1944
2.6397	2.0954	175000	8.8802	2.1838	0.0032	64.8308	4.1718	1.2146	0.1944
2.6608	2.1049	180000	8.8237	2.1774	0.0032	64.0079	4.1590	1.2146	0.1944
2.6339	2.1144	185000	8.7756	2.1720	0.0032	63.3068	4.1480	1.2146	0.1944
2.6327	3.0048	190000	8.7198	2.1656	0.0032	62.5058	4.1353	1.2146	0.1944
2.6281	3.0143	195000	8.6693	2.1598	0.0032	61.7858	4.1237	1.2146	0.1944
2.6281	3.0238	200000	8.6218	2.1543	0.0032	61.1157	4.1128	1.2146	0.1944
2.6206	3.0334	205000	8.5813	2.1496	0.0032	60.5393	4.1033	1.2146	0.1944
2.6253	3.0429	210000	8.5365	2.1443	0.0032	59.9116	4.0929	1.2146	0.1944
2.6222	3.0525	215000	8.4976	2.1398	0.0032	59.3662	4.0837	1.2146	0.1944
2.619	3.0620	220000	8.4646	2.1359	0.0032	58.9064	4.0759	1.2146	0.1944
2.6154	3.0715	225000	8.4291	2.1317	0.0032	58.4114	4.0675	1.2146	0.1944
2.6194	3.0811	230000	8.3955	2.1277	0.0032	57.9497	4.0596	1.2146	0.1944
2.6071	3.0906	235000	8.3591	2.1233	0.0032	57.4509	4.0509	1.2146	0.1944
2.6073	3.1001	240000	8.3282	2.1196	0.0032	57.0281	4.0435	1.2146	0.1944
2.6069	3.1097	245000	8.3028	2.1166	0.0032	56.6777	4.0374	1.2146	0.1944
2.5963	3.1192	250000	8.2717	2.1128	0.0032	56.2549	4.0299	1.2146	0.1944
2.5939	4.0095	255000	8.2485	2.1100	0.0032	55.9397	4.0243	1.2146	0.1944
2.6052	4.0191	260000	8.2190	2.1065	0.0032	55.5436	4.0172	1.2146	0.1944
2.6206	4.0286	265000	8.1968	2.1037	0.0032	55.2421	4.0117	1.2146	0.1944
2.5901	4.0381	270000	8.1721	2.1007	0.0032	54.9110	4.0057	1.2146	0.1944
2.6189	4.0477	275000	8.1531	2.0984	0.0032	54.6544	4.0010	1.2146	0.1944
2.6123	4.0572	280000	8.1302	2.0956	0.0032	54.3509	3.9955	1.2146	0.1944
2.5939	4.0668	285000	8.1105	2.0932	0.0032	54.0872	3.9906	1.2146	0.1944
2.5931	4.0763	290000	8.0902	2.0907	0.0032	53.8164	3.9856	1.2146	0.1944
2.5801	4.0858	295000	8.0712	2.0883	0.0032	53.5642	3.9809	1.2146	0.1944
2.5703	4.0954	300000	8.0520	2.0859	0.0032	53.3124	3.9762	1.2146	0.1944
2.5951	4.1049	305000	8.0361	2.0839	0.0032	53.1002	3.9722	1.2146	0.1944
2.5706	4.1144	310000	8.0240	2.0824	0.0032	52.9377	3.9691	1.2146	0.1944
2.5721	5.0048	315000	8.0057	2.0802	0.0032	52.6975	3.9646	1.2146	0.1944
2.5681	5.0143	320000	7.9894	2.0781	0.0032	52.4840	3.9605	1.2146	0.1944
2.5713	5.0238	325000	7.9746	2.0763	0.0032	52.2925	3.9569	1.2146	0.1944
2.5678	5.0334	330000	7.9619	2.0747	0.0032	52.1243	3.9536	1.2146	0.1944
2.5759	5.0429	335000	7.9467	2.0728	0.0032	51.9275	3.9498	1.2146	0.1944
2.5734	5.0525	340000	7.9344	2.0712	0.0032	51.7658	3.9467	1.2146	0.1944
2.5723	5.0620	345000	7.9246	2.0700	0.0032	51.6367	3.9442	1.2146	0.1944
2.5716	5.0715	350000	7.9128	2.0685	0.0032	51.4820	3.9412	1.2146	0.1944
2.5788	5.0811	355000	7.9013	2.0670	0.0032	51.3336	3.9383	1.2146	0.1944
2.5638	5.0906	360000	7.8881	2.0654	0.0032	51.1649	3.9351	1.2146	0.1944
2.5657	5.1001	365000	7.8774	2.0640	0.0032	51.0266	3.9323	1.2146	0.1944
2.5697	5.1097	370000	7.8698	2.0630	0.0032	50.9255	3.9304	1.2146	0.1944
2.5598	5.1192	375000	7.8584	2.0616	0.0032	50.7796	3.9275	1.2146	0.1944
2.5571	6.0095	380000	7.8515	2.0607	0.0032	50.6889	3.9257	1.2146	0.1944
2.5717	6.0191	385000	7.8408	2.0593	0.0032	50.5542	3.9230	1.2146	0.1944
2.5886	6.0286	390000	7.8331	2.0584	0.0032	50.4543	3.9211	1.2146	0.1944
2.5579	6.0381	395000	7.8248	2.0573	0.0032	50.3467	3.9189	1.2146	0.1944
2.5885	6.0477	400000	7.8188	2.0565	0.0032	50.2690	3.9174	1.2146	0.1944
2.584	6.0572	405000	7.8107	2.0555	0.0032	50.1654	3.9153	1.2146	0.1944
2.5663	6.0668	410000	7.8039	2.0546	0.0032	50.0788	3.9136	1.2146	0.1944
2.5658	6.0763	415000	7.7975	2.0538	0.0032	49.9948	3.9119	1.2146	0.1944
2.5549	6.0858	420000	7.7909	2.0530	0.0032	49.9114	3.9102	1.2146	0.1944
2.5445	6.0954	425000	7.7842	2.0521	0.0032	49.8271	3.9086	1.2146	0.1944
2.5732	6.1049	430000	7.7799	2.0515	0.0032	49.7709	3.9074	1.2146	0.1944
2.5483	6.1144	435000	7.7769	2.0512	0.0032	49.7301	3.9066	1.2146	0.1944
2.5494	7.0048	440000	7.7702	2.0503	0.0032	49.6461	3.9049	1.2146	0.1944
2.5482	7.0143	445000	7.7655	2.0497	0.0032	49.5863	3.9037	1.2146	0.1944
2.5514	7.0238	450000	7.7611	2.0491	0.0032	49.5319	3.9026	1.2146	0.1944
2.549	7.0334	455000	7.7576	2.0487	0.0032	49.4864	3.9017	1.2146	0.1944
2.5567	7.0429	460000	7.7537	2.0482	0.0032	49.4372	3.9007	1.2146	0.1944
2.5555	7.0525	465000	7.7504	2.0477	0.0032	49.3947	3.8998	1.2146	0.1944
2.5564	7.0620	470000	7.7482	2.0475	0.0032	49.3660	3.8993	1.2146	0.1944
2.5542	7.0715	475000	7.7453	2.0471	0.0032	49.3283	3.8985	1.2146	0.1944
2.5627	7.0811	480000	7.7427	2.0468	0.0032	49.2958	3.8978	1.2146	0.1944
2.5511	7.0906	485000	7.7396	2.0463	0.0032	49.2562	3.8970	1.2146	0.1944
2.5533	7.1001	490000	7.7372	2.0460	0.0032	49.2271	3.8964	1.2146	0.1944
2.557	7.1097	495000	7.7363	2.0459	0.0032	49.2147	3.8962	1.2146	0.1944
2.5467	7.1192	500000	7.7347	2.0457	0.0032	49.1938	3.8958	1.2146	0.1944
2.5482	8.0095	505000	7.7332	2.0455	0.0032	49.1749	3.8954	1.2146	0.1944
2.5599	8.0191	510000	7.7323	2.0454	0.0032	49.1640	3.8952	1.2146	0.1944
2.5788	8.0286	515000	7.7318	2.0453	0.0032	49.1576	3.8950	1.2146	0.1944
2.5479	8.0381	520000	7.7314	2.0453	0.0032	49.1523	3.8949	1.2146	0.1944

Framework versions

Transformers 4.57.0.dev0
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

Task: Feature Extraction
Library: transformers
Type: t5la_adapter
Created: Oct 28, 2025
Updated: Jan 16, 2026