Grimulkan

#5 opened 22 days ago by

the-hir0

New activity in grimulkan/lzlv-longLORA-70b-rope8-32k-6bpw-h8-exl2 2 months ago

EXL quantization at 4.0bpw?

3

#1 opened 3 months ago by

lazydog22

New activity in grimulkan/aurelian-v0.5-70b-rope8-32K-fp16 3 months ago

Further fine-tuning?

2

#4 opened 3 months ago by

jspr

New activity in grimulkan/aurelian-v0.5-70b-rope8-32K-fp16 4 months ago

Comparison with 0.1

13

#3 opened 4 months ago by

New activity in grimulkan/lzlv-longLORA-70b-rope8-32k-fp16 4 months ago

Benchmarks!

3

#2 opened 4 months ago by

New activity in grimulkan/llama2_70b_longlora_fp16_32k_ROPE8 4 months ago

Benchs

8

#1 opened 4 months ago by

Nexesenex

New activity in Nexesenex/MIstral-QUantized-70b_Miqu-1-70b-iMat.GGUF 4 months ago

Arch speculations

6

#3 opened 4 months ago by

New activity in grimulkan/Goliath-longLORA-120b-rope8-32k-fp16 4 months ago

More quants

#2 opened 4 months ago by

aikitoria

Observations+benchmarks

#1 opened 4 months ago by

New activity in grimulkan/aurelian-v0.5-70b-rope8-32K-fp16 4 months ago

Merging advice?

16

#2 opened 4 months ago by

sophosympatheia

New activity in grimulkan/aurelian-alpha0.1-70b-rope8-32K-fp16 4 months ago

2 bits GGUF SOTA quants?

31

#2 opened 5 months ago by

Nexesenex

New activity in grimulkan/lzlv-longLORA-70b-rope8-32k-fp16 4 months ago

Quantization calibration dataset

2

#1 opened 4 months ago by

Amajiro

New activity in grimulkan/aurelian-v0.5-70b-rope8-32K-fp16 4 months ago

Devolution into semi-nonsensical long words

6

#1 opened 4 months ago by

jspr

New activity in Squish42/bluemoon-fandom-1-1-rp-cleaned 5 months ago

Feedback and collaboration.

37

#1 opened about 1 year ago by

Squish42

New activity in grimulkan/aurelian-alpha0.1-70b-rope8-32K-fp16 5 months ago

You made gold!

3

#1 opened 5 months ago by

New activity in grimulkan/aurelian-alpha0.1-70b-rope8-32K-6bpw_h8_exl2 5 months ago

Repetition issue

#1 opened 6 months ago by

lazyDataScientist

New activity in Noeda/aurelian-alpha0.1-70b-rope8-32K-GGUF 6 months ago

GGUF parameters suggestion

#1 opened 6 months ago by

lazyDataScientist

New activity in jondurbin/airoboros-l2-70b-3.1.2 6 months ago

Space after [/INST]

7

#2 opened 7 months ago by

Satya93

New activity in Yukang/Llama-2-70b-longlora-32k 7 months ago

Rope scaling value?

#2 opened 7 months ago by

jurassicpark

New activity in jondurbin/airoboros-l2-70b-3.1.2 7 months ago

Ability to generalise

6

#1 opened 7 months ago by

vmajor

New activity in Yukang/LongAlpaca-70B 8 months ago

Any chance of a 16K model?

2

#2 opened 8 months ago by

MB7977

New activity in v2ray/jannie-log 8 months ago

Originals?

#1 opened 8 months ago by

Paper • 2309.12307 • Published Sep 21, 2023 • 82 •

commented a paper 8 months ago

LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models

8

New activity in Yukang/Llama-2-7b-longlora-100k-ft 8 months ago

Can't load the model using pipeline or tokenizer/model approach

#1 opened 8 months ago by

AhmadHakami

New activity in Yukang/Llama-2-13b-chat-longlora-32k-sft 8 months ago

Evaluation of long sequence of conversation

5

#1 opened 8 months ago by

cooee-ashutosh

New activity in Yukang/Llama-2-70b-longlora-32k 8 months ago

Training VRAM for 70B 32K

#1 opened 8 months ago by

New activity in Chat-Error/fiction.live-Kimiko-V2-70B 9 months ago

Share dataset?

#3 opened 9 months ago by