finetune-longt5

silmi224/finetune-longt5

finetune-longt5 is a summarization model from silmi224. Use it when you need a shorter version of a longer text. It is set up for transformers.

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

Downloads · 30 days

4% of all-time downloads

All-time downloads

294

Public

Parameters

248M

991 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors990 MB · 100%

At a glance

Task: Summarization
Library: transformers
Model type: longt5
Access: Public
Created: Jun 25, 2024
Updated: Jun 25, 2024
SHA: 8faa047f

Try a prompt

Model card

From the Hugging Face model README

finetune-longt5

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.9551
Rouge1 Precision: 0.2602
Rouge1 Recall: 0.3322
Rouge1 Fmeasure: 0.2861

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: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1 Fmeasure	Rouge1 Precision	Rouge1 Recall
2.5865	0.03	10	2.4019	0.2349	0.2481	0.2353
2.3384	0.06	20	2.2294	0.2404	0.2571	0.2385
2.0876	0.1	30	2.2887	0.2467	0.2653	0.2432
2.1205	0.13	40	2.2194	0.2517	0.2731	0.246
2.0637	0.16	50	2.2172	0.2577	0.2868	0.246
1.8623	0.19	60	2.2273	0.2613	0.2903	0.2497
0.0	0.22	70	nan	0.262	0.2911	0.25
1.8591	0.26	80	2.1895	0.2604	0.2901	0.2481
1.9819	0.29	90	2.1492	0.2663	0.2917	0.2575
1.9326	0.32	100	2.1248	0.2698	0.2964	0.2608
1.888	0.35	110	2.1253	0.2698	0.2954	0.2614
1.8934	0.38	120	2.0993	0.2705	0.3006	0.2589
1.8357	0.42	130	2.1050	0.2744	0.3001	0.267
1.8061	0.45	140	2.0705	0.2787	0.2924	0.2829
1.8444	0.48	150	2.1156	0.2739	0.2895	0.2762
1.83	0.51	160	2.0636	0.2773	0.2895	0.2831
1.7433	0.54	170	2.0857	0.2767	0.2811	0.29
1.7523	0.58	180	2.0809	0.2766	0.2798	0.2913
1.764	0.61	190	2.0351	0.2799	0.2751	0.3041
1.8441	0.64	200	2.0460	0.2804	0.2772	0.3025
1.7118	0.67	210	2.0319	0.2798	0.2767	0.3024
1.818	0.7	220	2.0287	0.2823	0.2666	0.318
1.7484	0.74	230	2.0084	0.2822	0.2653	0.3188
1.8224	0.77	240	2.0372	0.2787	0.2634	0.3132
1.6475	0.4	250	2.0281	0.2768	0.2594	0.3141
1.6915	0.42	260	1.9941	0.2851	0.2637	0.3269
1.7106	0.43	270	2.0204	0.2862	0.2718	0.3198
1.6382	0.45	280	2.0073	0.288	0.2657	0.3307
1.6017	0.46	290	2.0242	0.2847	0.2584	0.3328
1.8248	0.48	300	1.9996	0.2821	0.2595	0.3249
1.687	0.5	310	1.9801	0.2857	0.2664	0.3243
1.6993	0.51	320	1.9979	0.2837	0.2608	0.3271
1.6191	0.53	330	2.0025	0.285	0.2618	0.329
1.5409	0.54	340	1.9968	0.2851	0.2616	0.3293
1.6279	0.56	350	1.9940	0.2824	0.2601	0.3259
1.7538	0.58	360	1.9907	0.2803	0.2574	0.3234
1.6781	1.19	370	1.9684	0.2805	0.2565	0.3249
1.6106	1.22	380	1.9798	0.2842	0.2584	0.3314
1.5798	1.25	390	1.9940	0.2842	0.2581	0.3321
1.7134	1.28	400	1.9634	0.2851	0.2596	0.3313
1.6301	1.31	410	1.9644	0.2866	0.2606	0.3342
1.6657	1.35	420	1.9775	0.2861	0.2604	0.3325
1.7028	1.38	430	1.9792	0.2838	0.26	0.3275
1.7063	1.41	440	1.9728	0.2838	0.2598	0.3279
1.5825	1.44	450	1.9644	0.2838	0.259	0.3291
1.5955	1.47	460	1.9674	0.2864	0.2605	0.3329
1.6774	1.51	470	1.9718	0.288	0.2616	0.3354
1.7171	1.54	480	1.9602	0.2882	0.2619	0.3358
1.6631	1.57	490	1.9592	0.2873	0.2614	0.3335
1.5405	1.6	500	1.9625	0.2868	0.261	0.3329
1.5711	1.63	510	1.9690	0.2872	0.2614	0.3337
1.7382	1.67	520	1.9669	0.2873	0.262	0.3326
1.6345	1.7	530	1.9564	0.2867	0.2615	0.3323
1.6289	1.73	540	1.9558	0.2856	0.2604	0.3309
1.575	1.76	550	1.9620	0.2872	0.2616	0.333
1.682	1.79	560	1.9613	0.287	0.2615	0.3326
1.5385	1.83	570	1.9616	0.2869	0.2614	0.3324
1.62	1.86	580	1.9603	0.2868	0.2611	0.3325
1.6026	1.89	590	1.9589	0.2867	0.2611	0.3325
1.6511	1.92	600	1.9575	0.2865	0.2608	0.3322
1.5688	1.95	610	1.9555	0.2866	0.2608	0.3327
1.6092	1.99	620	1.9553	0.2863	0.2604	0.3324

Framework versions

Transformers 4.35.2
Pytorch 2.1.1+cu121
Datasets 2.14.5
Tokenizers 0.15.1

Task: Summarization
Library: transformers
Type: longt5
Created: Jun 25, 2024
Updated: Jun 25, 2024