Skip to content

Add utility functions and update training setup for minillm#1

Open
ghost wants to merge 19 commits into
expfrom
master
Open

Add utility functions and update training setup for minillm#1
ghost wants to merge 19 commits into
expfrom
master

Conversation

@ghost

@ghost ghost commented Jun 22, 2026

Copy link
Copy Markdown

No description provided.

Eamon2009 added 19 commits May 31, 2026 18:50
Add utility functions for LibTorch and custom Tensor conversion
Updated README to reflect new project name and details.
* removed: merge conflict in master and exp

* Delete main.cpp

* Add character-level DataLoader for training and validation batches

* config): add training and model hyperparameters for GPT architecture

* refactor: upgrade DataLoader from character-level to Byte Pair Encoding (BPE)

* model: add core architecture, cross-entropy loss, and AdamW optimizer

* inference: add inference SamplerParams and llama.cpp-style repetition penalty

* Delete .github directory

* main:  training, evaluation loops, and interactive chat CLI

* Delete CUDA directory

* Update requirements.txt

* Add datasets to requirements pip

* Add wandb and numpy to requirements

* Revise README for minillm project introduction

Updated README to reflect new project name and details.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant