vishwr/llama-3.2-1b-resume-assistant-lora-v1
Model Details
- Artifact kind:
adapter - Base model:
mlx-community/Llama-3.2-1B-Instruct-bf16 - Training method:
LoRA SFT with prompt masking - Intended runtime:
MLX / MLX-LM, with optional post-training Ollama import
Intended Use
This model is intended for resume-assistant style tasks such as summarization, critique, category classification, skills extraction, and rewrite assistance.
Dataset Provenance
- Source rows: 2988
- Kept rows: 1384
- Max sequence length: 1024
- Split counts: {"train": 1242, "valid": 66, "test": 76}
Evaluation Summary
{
"sample_count": 15,
"family_count": 9,
"adapter_avg_rouge_l_f1": 0.0,
"base_avg_rouge_l_f1": 0.0833
}
Limitations
- This model is fine-tuned on resume-oriented prompts and can overfit to that style.
- It should not be treated as a factual source for employment decisions.
- Contact details and work history can appear in training data; do not publish publicly without review.
Privacy
Keep this repository private unless privacy and licensing review are complete.
Hardware compatibility
Log In to add your hardware
Quantized
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for vishwr/llama-3.2-1b-resume-assistant-lora-v1
Base model
mlx-community/Llama-3.2-1B-Instruct-bf16
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir llama-3.2-1b-resume-assistant-lora-v1 vishwr/llama-3.2-1b-resume-assistant-lora-v1