ruadapt_solar_10.7_part1

RefalMachine/ruadapt_solar_10.7_part1

ruadapt_solar_10.7_part1 is a text generation model from RefalMachine. Use it when you need the model to write or continue 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

578

Public

Parameters

10.7B

21.5 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors21.5 GB · 100%

At a glance

Task: Text Generation
Library: transformers
Model type: llama
Access: Public
Created: Feb 17, 2024
Updated: Feb 17, 2024
SHA: 39525baf

Try a prompt

Base models

upstage/SOLAR-10.7B-v1.0

Spaces

Model card

From the Hugging Face model README

solar_10.7_darulm_unigram_proj_init_8node_darulm_part1_v3_1.0_512_12_02_24

This model is a fine-tuned version of ../solar_darulm_unigram_proj_init_17_01_24 on the None dataset. It achieves the following results on the evaluation set:

Loss: 2.3397
Accuracy: 0.5164

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: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 16
gradient_accumulation_steps: 8
total_train_batch_size: 128
total_eval_batch_size: 16
optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
lr_scheduler_type: linear
num_epochs: 1.0
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.6722	0.01	500	2.4811	0.4951
2.6243	0.02	1000	2.4459	0.4999
2.6051	0.04	1500	2.4295	0.5025
2.5901	0.05	2000	2.4194	0.5037
2.5852	0.06	2500	2.4124	0.5049
2.5818	0.07	3000	2.4072	0.5054
2.5801	0.09	3500	2.4024	0.5059
2.5626	0.1	4000	2.3988	0.5070
2.5697	0.11	4500	2.3958	0.5073
2.5532	0.12	5000	2.3928	0.5079
2.5505	0.13	5500	2.3904	0.5080
2.5497	0.15	6000	2.3872	0.5086
2.5636	0.16	6500	2.3857	0.5089
2.5483	0.17	7000	2.3835	0.5092
2.5505	0.18	7500	2.3813	0.5097
2.5419	0.2	8000	2.3796	0.5096
2.5467	0.21	8500	2.3786	0.5099
2.5419	0.22	9000	2.3769	0.5102
2.5269	0.23	9500	2.3754	0.5105
2.5315	0.24	10000	2.3740	0.5106
2.5442	0.26	10500	2.3728	0.5108
2.5318	0.27	11000	2.3713	0.5112
2.5242	0.28	11500	2.3702	0.5113
2.5178	0.29	12000	2.3698	0.5112
2.5345	0.31	12500	2.3687	0.5114
2.531	0.32	13000	2.3675	0.5115
2.5304	0.33	13500	2.3661	0.5118
2.5264	0.34	14000	2.3653	0.5121
2.5281	0.35	14500	2.3647	0.5123
2.5259	0.37	15000	2.3636	0.5123
2.5075	0.38	15500	2.3629	0.5122
2.5147	0.39	16000	2.3621	0.5127
2.5137	0.4	16500	2.3611	0.5128
2.5206	0.42	17000	2.3603	0.5129
2.5153	0.43	17500	2.3597	0.5128
2.5184	0.44	18000	2.3590	0.5130
2.5104	0.45	18500	2.3581	0.5132
2.5085	0.46	19000	2.3577	0.5134
2.509	0.48	19500	2.3572	0.5135
2.5143	0.49	20000	2.3564	0.5135
2.5124	0.5	20500	2.3555	0.5137
2.5107	0.51	21000	2.3546	0.5139
2.5034	0.53	21500	2.3543	0.5140
2.4922	0.54	22000	2.3538	0.5139
2.514	0.55	22500	2.3532	0.5140
2.5199	0.56	23000	2.3527	0.5141
2.4926	0.57	23500	2.3521	0.5142
2.5104	0.59	24000	2.3517	0.5142
2.5067	0.6	24500	2.3511	0.5144
2.5055	0.61	25000	2.3508	0.5142
2.5011	0.62	25500	2.3502	0.5146
2.4931	0.64	26000	2.3496	0.5147
2.4965	0.65	26500	2.3491	0.5147
2.495	0.66	27000	2.3488	0.5146
2.5051	0.67	27500	2.3481	0.5150
2.51	0.68	28000	2.3478	0.5150
2.4883	0.7	28500	2.3474	0.5152
2.4973	0.71	29000	2.3470	0.5151
2.4939	0.72	29500	2.3464	0.5153
2.4952	0.73	30000	2.3461	0.5153
2.5028	0.75	30500	2.3459	0.5154
2.4979	0.76	31000	2.3454	0.5154
2.4928	0.77	31500	2.3450	0.5155
2.501	0.78	32000	2.3446	0.5156
2.5	0.79	32500	2.3443	0.5156
2.4865	0.81	33000	2.3438	0.5156
2.4898	0.82	33500	2.3434	0.5157
2.4977	0.83	34000	2.3430	0.5160
2.4904	0.84	34500	2.3427	0.5157
2.4779	0.86	35000	2.3424	0.5159
2.4792	0.87	35500	2.3420	0.5159
2.4931	0.88	36000	2.3419	0.5160
2.4997	0.89	36500	2.3416	0.5160
2.4986	0.9	37000	2.3414	0.5161
2.4965	0.92	37500	2.3411	0.5162
2.4743	0.93	38000	2.3409	0.5162
2.497	0.94	38500	2.3406	0.5163
2.4942	0.95	39000	2.3404	0.5162
2.4907	0.97	39500	2.3402	0.5163
2.4821	0.98	40000	2.3400	0.5163
2.4857	0.99	40500	2.3398	0.5163

Framework versions

Transformers 4.37.2
Pytorch 2.1.2
Datasets 2.16.1
Tokenizers 0.15.2

Task: Text Generation
Library: transformers
Type: llama
Created: Feb 17, 2024
Updated: Feb 17, 2024