microsoft/phi-2 · Discussions

gave a bibtex citation

#133 opened about 2 months ago by

squidWorm

Phi 2 Instruct: an instruction following Phi 2 SLM that has undergone SFT and DPO

#132 opened 3 months ago by

rasyosef

Update README.md

#131 opened 3 months ago by

Yaserrati

# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-2") model = AutoModelForCausalLM.from_pretrained("microsoft/phi-2")

#130 opened 4 months ago by

Khaliq88

Tokenizer.model is not present in Files and versions

#129 opened 5 months ago by

naina1S

max length control

#127 opened 6 months ago by

Ashvithaa

Minumum gpu ram to fine tune phi-2 on custom data and what should be the datarows to fine tune phi-2 model.

#126 opened 7 months ago by

Ayushmzn

Model inference speed is too slow (positively related to max_new_tokens length)

#125 opened 7 months ago by

Kyrieeee

inference Endpoints error

#124 opened 8 months ago by

gawon16

Generation after finetuning does not ends at EOS token

1

#123 opened 8 months ago by

zokica

Trouble with Phi2 Tokenisation

1

#116 opened 9 months ago by

riedgar-ms

Target modules {'out_proj', 'Wqkv'} is not found in the phi-2 model how can I fix this error?

2

#115 opened 9 months ago by

roy1109

Fine tuned phi2 model loses context once loaded from local

1

#112 opened 9 months ago by

tatvamasi

Disabled autocast

9

#109 opened 9 months ago by

miguelcarv

Model not loading from local directory

1

#108 opened 9 months ago by

mdaniyal214

Phi-2 prompt engineering

1

#107 opened 10 months ago by

setareh1

Training model breaks with flash_attn_2. Error "NameError: name 'index_first_axis' is not defined"

12

#105 opened 10 months ago by

praveeny

I encountered an error when attempting to train phi

#104 opened 10 months ago by

PhelixZhen

instruct fone tune phi2 on custom data

1

#103 opened 10 months ago by

akashAD

How to finetune Phi2 using RoPE and QLoRA for long text summary generation?

4

#102 opened 10 months ago by

parikshit1619

supervised finetuning error

#96 opened 10 months ago by

AlexMercerXX

How to generate a gguf file from a fine tuned phi-2 ?

1

#93 opened 10 months ago by

fivetech

Loading a qlora finetuned checkpoint with the new updates outputs differently

6

#88 opened 10 months ago by

YaYaGeGe

How to reduce nonsense generation?

9

#81 opened 10 months ago by

Kelmeilia

Cannot load the model PhiForSequenceClassificationModified(basemodel.config)

3

#80 opened 10 months ago by

lyliiiii

Irrelevant text generation while prompting

5

#63 opened 11 months ago by

ad6398

run Phi-2 on your CPU

12

#62 opened 11 months ago by

J22

Responses by phi-2 are off, it's depressed and insults me for no reason whatsover?

7

#61 opened 11 months ago by

NightcoreSkies

"Instruct: <prompt>\nOutput:" or "Instruction: <prompt>\nOutput:"

5

#60 opened 11 months ago by

J22

What is the best way for the inference process in LORA in PEFT approach

#53 opened 11 months ago by

Pradeep1995

[Positivity] What is it good for?

1

#52 opened 11 months ago by

dagbs

Review on Phi-2

4

#45 opened 11 months ago by

tusharpaul

Can inference be done with the model converted to OpenVINO?

2

#44 opened 11 months ago by

DeltaLux

Unimpressed by Phi-2

6

#39 opened 11 months ago by

Aditiyadav

How to fine-tune this? + Training code

43

#19 opened 11 months ago by

cekal