You can deploy this model and start earning money today!

RichardErkhov/
thrunlab_-_relu_llama_7b_hf2_refined_web_relu_2024-03-27-4bits

Name: thrunlab_-_relu_llama_7b_hf2_refined_web_relu_2024-03-27-4bits
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

relu_llama_7b_hf2_refined_web_relu_2024-03-27 - bnb 4bits

Original model description:

base_model: meta-llama/Llama-2-7b-hf tags:

This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the None dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

Training Loss	Epoch	Step	Validation Loss
9.7497	0.0	25	9.2613
8.0241	0.01	50	7.7537
6.8496	0.01	75	6.7390
5.8173	0.02	100	5.6115
4.9513	0.02	125	4.8869
4.4377	0.02	150	4.3805
4.0562	0.03	175	3.9836
3.6569	0.03	200	3.7061
3.5457	0.04	225	3.5205
3.4229	0.04	250	3.3833
3.214	0.04	275	3.2813
3.1747	0.05	300	3.2035
3.1654	0.05	325	3.1411
2.8801	0.06	350	3.0881
3.0155	0.06	375	3.0453
3.1558	0.06	400	3.0078
3.0349	0.07	425	2.9774
2.9819	0.07	450	2.9491
2.8286	0.08	475	2.9239
2.8718	0.08	500	2.9013
2.9262	0.08	525	2.8812
2.8091	0.09	550	2.8623
2.8676	0.09	575	2.8440
2.7304	0.1	600	2.8292
2.8206	0.1	625	2.8158
2.8212	0.1	650	2.8037
2.8385	0.11	675	2.7907
2.7437	0.11	700	2.7797
2.7773	0.12	725	2.7696
2.6785	0.12	750	2.7611
2.7582	0.12	775	2.7510
2.7785	0.13	800	2.7414
2.7549	0.13	825	2.7339
2.7228	0.14	850	2.7257
2.5928	0.14	875	2.7189
2.7048	0.14	900	2.7118
2.6131	0.15	925	2.7052
2.7515	0.15	950	2.6994
2.7365	0.16	975	2.6933
2.7635	0.16	1000	2.6882
2.7883	0.16	1025	2.6841
2.7032	0.17	1050	2.6782
2.714	0.17	1075	2.6728
2.6427	0.18	1100	2.6684
2.6727	0.18	1125	2.6644
2.7536	0.18	1150	2.6593
2.7379	0.19	1175	2.6547
2.5601	0.19	1200	2.6500
2.6281	0.2	1225	2.6461
2.6526	0.2	1250	2.6421
2.724	0.2	1275	2.6388
2.6527	0.21	1300	2.6347
2.5999	0.21	1325	2.6305
2.525	0.22	1350	2.6270