README.md

metadata

tags:
  - generated_from_trainer
metrics:
  - bleu
model-index:
  - name: es_fi_all_quy
    results: []

es_fi_all_quy

This model is a fine-tuned version of nouman-10/es_fi_all_quy on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4691
• Bleu: 1.3097
• Chrf: 33.573
• Gen Len: 42.0221

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Chrf	Gen Len
0.2653	0.09	1000	0.4870	1.3376	32.1158	43.001
0.2668	0.17	2000	0.4826	1.3753	32.002	46.505
0.2567	0.26	3000	0.4820	1.2717	31.9404	46.7274
0.2561	0.34	4000	0.4825	1.4256	32.4758	41.7274
0.2618	0.43	5000	0.4850	1.6935	33.2306	37.2012
0.2705	0.51	6000	0.4723	1.372	32.4431	46.84
0.2681	0.6	7000	0.4758	1.4419	32.8507	45.6016
0.2629	0.68	8000	0.4737	1.4636	33.3288	40.0382
0.2773	0.77	9000	0.4715	1.2296	33.1241	41.502
0.2702	0.85	10000	0.4663	1.2579	32.8273	44.9034
0.2683	0.94	11000	0.4694	1.6207	32.8479	42.3964
0.259	1.02	12000	0.4766	1.4934	32.6413	41.0815
0.2537	1.11	13000	0.4713	1.7586	33.3814	39.9638
0.2516	1.19	14000	0.4724	1.593	33.4105	41.832
0.2574	1.28	15000	0.4749	1.3373	33.3664	42.3662
0.2523	1.37	16000	0.4701	1.1924	32.6157	42.7706
0.2462	1.45	17000	0.4710	1.5688	33.5992	40.5282
0.2513	1.54	18000	0.4723	1.2722	32.1578	47.4225
0.2504	1.62	19000	0.4728	1.3897	32.6709	40.8893
0.2502	1.71	20000	0.4714	1.5999	33.6673	41.5362
0.2434	1.79	21000	0.4715	1.9393	33.6971	40.8944
0.2483	1.88	22000	0.4688	1.8308	34.1117	37.7565
0.2435	1.96	23000	0.4693	1.8643	34.5409	38.6237
0.2377	2.05	24000	0.4702	1.6217	33.6401	40.4779
0.235	2.13	25000	0.4707	1.5441	33.588	39.8974
0.2345	2.22	26000	0.4710	2.0248	33.7469	37.2535
0.2423	2.3	27000	0.4691	1.9699	33.4757	37.9889
0.2388	2.39	28000	0.4669	1.5651	33.1965	39.7646
0.2367	2.47	29000	0.4682	1.69	33.9955	38.3199
0.2392	2.56	30000	0.4720	1.9972	33.902	41.2525
0.2382	2.65	31000	0.4721	2.0682	33.6693	38.3833
0.2373	2.73	32000	0.4690	2.0952	33.553	38.3229
0.2356	2.82	33000	0.4691	1.3097	33.573	42.0221

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu117
• Datasets 2.11.0
• Tokenizers 0.13.3