Abhi Venigalla

abhi-mosaic's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes

New activity in mosaicml/mpt-30b-instruct over 1 year ago

minor prose tweaks

#1 opened over 1 year ago by dblalock

New activity in mosaicml/mpt-30b over 1 year ago

prose tweaks

#2 opened over 1 year ago by dblalock

New activity in mosaicml/mpt-7b-chat over 1 year ago

there are something wrong when the latest code generate longer text

4

#19 opened over 1 year ago by lvkaokao

New activity in mosaicml/mpt-7b over 1 year ago

MPT-7b on colab - RAM of GPU not used

5

#50 opened over 1 year ago by vi-c

Pad_token_id of MPT-7B

2

#49 opened over 1 year ago by Trung-Dung

Running on single Nvidia K80 GPU with large context to generate long output

3

#29 opened over 1 year ago by airtable

Speed on CPU

13

#8 opened over 1 year ago by zokica

H100 TransformerEngine

3

#14 opened over 1 year ago by SinanAkkoyun

Using `self.transformer.wte.weight` directly for LM head breaks HF accelerate device map auto infer on multi-gpu

3

#46 opened over 1 year ago by shijie-wu

Create requirements.txt

2

#22 opened over 1 year ago by shijie-wu

New activity in mosaicml/mpt-7b-instruct over 1 year ago

configuration / penalty to lower repetition?

5

#32 opened over 1 year ago by mfab

How much GPU memory is needed to finetune MPT-7B Instruct model?

2

#31 opened over 1 year ago by skshreyas714

Provide fine-tuning example notebook using hf transformers

6

#22 opened over 1 year ago by MakerMotion

GPU Memory / RAM requierements

8

#19 opened over 1 year ago by Rbn3D

Help Needed!! Text Generation Taking Too Long

2

#17 opened over 1 year ago by debajyoti111

Error in Triton implementation

2

#9 opened over 1 year ago by narenzen

Error in Installation Setup

2

#24 opened over 1 year ago by souvik0306

KeyError in triton implementation

7

#25 opened over 1 year ago by datacow

8bit and sharded weights

2

#37 opened over 1 year ago by ThreeBlessings

ValueError: MPTForCausalLM does not support `device_map='auto'` yet.

2

#38 opened over 1 year ago by AayushShah