bart-large-cnndm

ryusangwon/bart-large-cnndm

bart-large-cnndm is a machine learning model from ryusangwon. 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 transformers. The card lists the license as apache-2.0.

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

Downloads · 30 days

18% of all-time downloads

All-time downloads

Public

Parameters

406M

2.2 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors3.8 GB · 100%

At a glance

Library: transformers
License: apache-2.0
Model type: bart
Access: Public
Created: Feb 2, 2024
Updated: Feb 7, 2024
SHA: 7b5e4ff6

Base models

facebook/bart-large

Model card

From the Hugging Face model README

cnn_dailymail_726_bart-large

This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.8412
Rouge1: 0.2469
Rouge2: 0.1266
Rougel: 0.2074
Rougelsum: 0.2332
Gen Len: 20.0

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: 16
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.9706	0.22	500	0.9015	0.237	0.1181	0.1979	0.2232	19.9999
0.9212	0.45	1000	0.8771	0.237	0.1193	0.199	0.2233	20.0
0.8991	0.67	1500	0.8572	0.2443	0.1238	0.2045	0.2304	20.0
0.9085	0.89	2000	0.8519	0.2404	0.1227	0.2022	0.2269	20.0
0.8494	1.11	2500	0.8471	0.2437	0.1233	0.2041	0.2298	20.0
0.832	1.34	3000	0.8400	0.2438	0.1248	0.2055	0.2301	20.0
0.8522	1.56	3500	0.8393	0.2417	0.1242	0.2043	0.2283	20.0
0.8494	1.78	4000	0.8338	0.2436	0.1239	0.2047	0.23	19.9999
0.7729	2.01	4500	0.8332	0.2431	0.1253	0.2048	0.2298	20.0
0.7761	2.23	5000	0.8323	0.2477	0.1264	0.207	0.2335	19.9994
0.7788	2.45	5500	0.8277	0.2473	0.1259	0.2068	0.2333	20.0
0.7832	2.67	6000	0.8251	0.2453	0.126	0.2061	0.2317	20.0
0.7888	2.9	6500	0.8239	0.242	0.1241	0.2037	0.2287	20.0
0.7413	3.12	7000	0.8360	0.2394	0.1228	0.2017	0.2258	20.0
0.7438	3.34	7500	0.8283	0.2462	0.1267	0.2072	0.2326	19.9999
0.7271	3.57	8000	0.8275	0.2406	0.1235	0.2028	0.2276	20.0
0.7435	3.79	8500	0.8221	0.2451	0.1254	0.2055	0.2311	19.9998
0.7072	4.01	9000	0.8277	0.2437	0.1251	0.2049	0.2301	19.9999
0.708	4.24	9500	0.8270	0.2465	0.1263	0.2067	0.2325	19.9999
0.7058	4.46	10000	0.8279	0.2424	0.1249	0.2045	0.229	19.9999
0.6918	4.68	10500	0.8248	0.246	0.1259	0.2063	0.232	19.9998
0.7121	4.9	11000	0.8231	0.2457	0.126	0.2058	0.232	19.9999
0.6667	5.13	11500	0.8297	0.2458	0.1262	0.2066	0.2323	19.9996
0.6767	5.35	12000	0.8309	0.2469	0.1269	0.2071	0.2332	19.9996
0.6961	5.57	12500	0.8299	0.247	0.1271	0.2074	0.2333	20.0
0.6842	5.8	13000	0.8333	0.2473	0.127	0.2077	0.2336	19.9996
0.6485	6.02	13500	0.8360	0.2454	0.1259	0.2061	0.2316	19.9998
0.6651	6.24	14000	0.8349	0.2454	0.126	0.2062	0.2314	20.0
0.6483	6.46	14500	0.8331	0.2454	0.1258	0.2058	0.2316	20.0
0.6626	6.69	15000	0.8309	0.2468	0.127	0.2069	0.2328	19.9996
0.6675	6.91	15500	0.8337	0.2448	0.1255	0.2056	0.231	19.9999
0.6479	7.13	16000	0.8387	0.2471	0.1267	0.2074	0.2333	19.9999
0.6506	7.36	16500	0.8377	0.2474	0.1264	0.2071	0.2335	19.9999
0.643	7.58	17000	0.8369	0.2454	0.1259	0.2059	0.2318	20.0
0.6262	7.8	17500	0.8378	0.2466	0.1269	0.2071	0.233	19.9997
0.6235	8.02	18000	0.8415	0.2458	0.1266	0.2065	0.2321	20.0
0.6081	8.25	18500	0.8421	0.2465	0.1267	0.2069	0.2326	19.9997
0.6257	8.47	19000	0.8409	0.2477	0.1267	0.2075	0.2337	19.9999
0.6187	8.69	19500	0.8381	0.2459	0.1264	0.2066	0.2321	19.9997
0.6178	8.92	20000	0.8384	0.248	0.1273	0.2079	0.2339	19.9996
0.6018	9.14	20500	0.8432	0.2468	0.1265	0.2071	0.2329	20.0
0.6235	9.36	21000	0.8418	0.2469	0.1265	0.207	0.233	20.0
0.606	9.58	21500	0.8418	0.2464	0.1264	0.207	0.2327	19.9999
0.6016	9.81	22000	0.8412	0.2469	0.1266	0.2074	0.2332	20.0

Framework versions

Transformers 4.37.2
Pytorch 2.2.0+cu121
Datasets 2.16.1
Tokenizers 0.15.1

Library: transformers
Type: bart
License: apache-2.0
Created: Feb 2, 2024
Updated: Feb 7, 2024