Instructions to use rmtlabs/IMCatalina-v1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use rmtlabs/IMCatalina-v1.0 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="rmtlabs/IMCatalina-v1.0")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("rmtlabs/IMCatalina-v1.0")
model = AutoModelForCausalLM.from_pretrained("rmtlabs/IMCatalina-v1.0")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use rmtlabs/IMCatalina-v1.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "rmtlabs/IMCatalina-v1.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rmtlabs/IMCatalina-v1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/rmtlabs/IMCatalina-v1.0

SGLang

How to use rmtlabs/IMCatalina-v1.0 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "rmtlabs/IMCatalina-v1.0" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rmtlabs/IMCatalina-v1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "rmtlabs/IMCatalina-v1.0" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rmtlabs/IMCatalina-v1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use rmtlabs/IMCatalina-v1.0 with Docker Model Runner:
```
docker model run hf.co/rmtlabs/IMCatalina-v1.0
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

IMCatalina-v1.0

Model summary

IMCatalina-v1.0 is a fully fine-tuned version of Phi-4 specialized in recruitment document processing.

The model focuses exclusively on:

Parsing unstructured CVs/resumes
Converting CV content into structured formats (JSON / YAML)
Generating professional job descriptions from structured inputs

This model was trained end-to-end (full fine-tuning) and does not perform candidate scoring, ranking, or hiring decisions.

Intended use

Primary use cases

CV and resume parsing
Structured CV normalization (JSON / YAML)
Extraction of skills, roles, education, and experience
Job description generation for recruitment platforms
Preprocessing for ATS and HR systems

Explicitly out-of-scope

Candidate ranking or scoring
Hiring recommendations
Candidate–job matching
Automated decision-making
Psychological or behavioral inference

Model details

Base model: microsoft/phi-4
Model type: Decoder-only causal language model
Architecture: Transformer (Phi family)
Parameters: ~14B
Context length: up to 16k tokens
Languages: English
Training type: Full fine-tuning

Training

Training data

Domain: Recruitment and HR documentation
Data type: Synthetic and curated structured data
Formats:
- Instruction–response
- Schema-constrained generation
Content includes:
- CVs and resumes
- Job descriptions
- Skills, roles, education, and experience fields
Data processing:
- Deduplication
- Schema validation
- Removal of malformed samples
- Consistency and format checks

No real personal data was intentionally included in the training datasets.

Downloads last month: 4

Safetensors

Model size

15B params

Tensor type

BF16

Model tree for rmtlabs/IMCatalina-v1.0

Base model

microsoft/phi-4

Finetuned

(100)

this model

Quantizations

1 model