allenai / Olmo-Hybrid-Instruct-DPO-7B

QUANTIZED BY: UnstableLlama

Information

⚠️ Requires ExLlamaV3 v0.0.26
exl3 quantizations of Olmo-Hybrid-Instruct-DPO-7B via exllamav3.
repo generated automatically with ezexl3.

Repo Data

REVISION	GiB	KL DIV	PPL
2.00bpw	2.55	0.2924	14.0737
2.10bpw	2.62	0.2610	13.6080
3.00bpw	3.33	0.0689	12.4490
3.10bpw	3.39	0.0613	12.3476
4.00bpw	4.10	0.0173	12.0738
5.00bpw	4.88	0.0053	12.0518
6.00bpw	5.65	0.0018	12.0559
8.00bpw	7.20	0.0007	12.0422
bf16	13.84	0.0000	12.0358

CLI Download

hf download UnstableLlama/Olmo-Hybrid-Instruct-DPO-7B-exl3 --revision "4.00bpw" --local-dir ./Olmo-Hybrid-Instruct-DPO-7B-exl3-4.00bpw

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for UnstableLlama/Olmo-Hybrid-Instruct-DPO-7B-exl3

Base model

Finetuned

Quantized

(1)

this model