Releases · microsoft/windows-ai-studio-templates

What's Changed

Remove DataParallel for preview version by @vriveras in #85
Adding base only inferencing by @vriveras in #87
Moving Phi-2 to HF provider by @vriveras in #89
Fixing true to True by @vriveras in #91
New style of prompt templates by @rodrigovaras in #101
Rename type to task by @kuojianlu in #102
feat:Adding Phi3 initial template for local fine-tuning by @vriveras in #103
feat: add inference => finetuning project template name mapping by @a1exwang in #104
fix:Fixing prompt template for Phi3 by @vriveras in #105
fix: Missed one end tag and removing a new line by @vriveras in #106
feat: Adding Phi3 inference models for 128k context window. by @vriveras in #108
Adding new DML Index for models while we have the official one. by @vriveras in #110
fix: remove huggingface secret in bicep by @XiaofuHuang in #111
feat: increase GPU instance number by @XiaofuHuang in #107
fix: update aca image by @swatDong in #112
feat: Add inference model metadata by @a1exwang in #109
fix: hf download and image by @swatDong in #113
feat: add remote support to all template by @XiaofuHuang in #115
fix: update replica timeout to 3 hours by @XiaofuHuang in #116
fix: move timeout to be a parameter by @XiaofuHuang in #117
fix: add default location in bicep parameters by @XiaofuHuang in #118
Update olive configs to match new format by @kaysabelle in #114
docs: add remote development into readme by @XiaofuHuang in #119
fix: for remote finetuning, add step to check finetune output by @swatDong in #121
fix: update resource prefix by @XiaofuHuang in #122
docs: add remote log to readme by @swatDong in #123
Add phi-3 fine-tuning mapping by @kuojianlu in #125
Adding temporary olive wheel until release by @vriveras in #124
docs: add catalog into README by @XiaofuHuang in #127
docs: update readme file with the latest UI changes and comments by @XiaofuHuang in #126
feat: adding cuda variant back for linux build by @vriveras in #128
docs: add private feature flag in remote development by @XiaofuHuang in #129
Add templates for Meta-Llama-3-8B by @kaysabelle in #120
Remove Olive wheels in favor of olive-ai version 0.6.0 by @kaysabelle in #131
Remove wandb and update transformers across all model templates by @kaysabelle in #132
fix: update remote config according to the latest deps version by @XiaofuHuang in #133
fix: specify inference dependency version in command by @XiaofuHuang in #134
fix: decrease startup probe period from 4 min to 2 min by @XiaofuHuang in #135
fix:moving to current official list of models by @vriveras in #136
fix: update phi-2 model version in remote development by @XiaofuHuang in #137
fix: change adapter path by @XiaofuHuang in #138
fix: add mistral model playground and finetune mapping by @XiaofuHuang in #139
fix: adapter path according to model by @XiaofuHuang in #140
feat:adding mistralai/Mistral-7B-Instruct-v0.2 template by @vriveras in #141
Update mistral model in finetuning command by @kaysabelle in #142
feat: add OpenAI compatible inference API by @a1exwang in #146
feat: add OpenAI compatible inference API for other templates by @a1exwang in #147
Release/v0.3.4 by @a1exwang in #148