Adi Renduchintala

adapter tuning for Megatron GPT models

1

# What does this PR do ? Add a one line overview of what this PR aims to accomplish. **Collection**: [Note which collection this PR will affect] # Changelog -...

prefix tuning for Megatron gpt models

14

# What does this PR do ? implements prefix tuning for Megatron GPT models. **Collection**: `examples/nlp/languge_modeling` # Changelog - Add specific line by line info of high level changes in...

IA3 support for GPT and T5

6

Signed-off-by: arendu # What does this PR do ? Add a one line overview of what this PR aims to accomplish. **Collection**: [Note which collection this PR will affect] #...

PEFT training of an LLM as an embedding model

4

# What does this PR do ? train a Lora model to perform sentence embedding tasks which can be used for IR. **Collection**: [NLP] # Changelog - Add specific line...

add alpha scaling to lora

2

# What does this PR do ? step in line with the original lora and added alpha. **Collection**: [NLP] # Changelog - Add specific line by line info of high...

# What does this PR do ? scripts to convert hugging face lora model into nemo. **Collection**: [Note which collection this PR will affect] # Changelog - Add specific line...

Currently, beir evaluates a model by calling its encode_queries and encoder_corpus functions. These in turn call the `forward` methods of the model. This works great for pytorch-based models, but not...

# What does this PR do ? allows the option to learn embeddings using matryoshka embedding learning. **Collection**: [Note which collection this PR will affect] # Changelog - Add specific...

# What does this PR do ? enables training of a contrastive reranker model based on GPT architecture and PEFT. Used for reranking the results of an embedding model. **Collection**:...

Mamba Refac

1

# What does this PR do ? incorporate mamba as a part of gpt models. **Collection**: [NLP] # Changelog - Add specific line by line info of high level changes...