Instructions to use TakalaWang/Discussion-Phi-4-multimodal-instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TakalaWang/Discussion-Phi-4-multimodal-instruct with Transformers:

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

pipe = pipeline("text-generation", model="TakalaWang/Discussion-Phi-4-multimodal-instruct", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("TakalaWang/Discussion-Phi-4-multimodal-instruct", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use TakalaWang/Discussion-Phi-4-multimodal-instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TakalaWang/Discussion-Phi-4-multimodal-instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TakalaWang/Discussion-Phi-4-multimodal-instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TakalaWang/Discussion-Phi-4-multimodal-instruct

SGLang

How to use TakalaWang/Discussion-Phi-4-multimodal-instruct 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 "TakalaWang/Discussion-Phi-4-multimodal-instruct" \
    --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": "TakalaWang/Discussion-Phi-4-multimodal-instruct",
		"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 "TakalaWang/Discussion-Phi-4-multimodal-instruct" \
        --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": "TakalaWang/Discussion-Phi-4-multimodal-instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use TakalaWang/Discussion-Phi-4-multimodal-instruct with Docker Model Runner:
```
docker model run hf.co/TakalaWang/Discussion-Phi-4-multimodal-instruct
```

Discussion-Phi-4-multimodal-instruct-w-asr

This model is a fine-tuned version of microsoft/Phi-4-multimodal-instruct on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 13.8976

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 4e-05
train_batch_size: 1
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 16
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-07 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 50
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
1.9923	0.2235	10	14.2728
0.3819	0.4469	20	14.0122
0.249	0.6704	30	14.0852
0.2951	0.8939	40	14.0115
0.2179	1.1117	50	14.0154
0.2839	1.3352	60	13.9879
0.2388	1.5587	70	14.0103
0.2687	1.7821	80	13.9336
0.1469	2.0	90	13.9879
0.1891	2.2235	100	13.9434
0.133	2.4469	110	14.0584
0.1759	2.6704	120	13.8902
0.1613	2.8939	130	13.8976

Framework versions

Transformers 4.51.3
Pytorch 2.4.1+cu124
Datasets 3.5.1
Tokenizers 0.21.1

Downloads last month: 7

Safetensors

Model size

6B params

Tensor type

F32

BF16

Model tree for TakalaWang/Discussion-Phi-4-multimodal-instruct

Base model

microsoft/Phi-4-multimodal-instruct

Finetuned

(49)

this model

Collection including TakalaWang/Discussion-Phi-4-multimodal-instruct

Discussion

Collection

11 items • Updated Mar 2