Text Generation
Transformers
Safetensors
PyTorch
English
rose_x1
causal-lm
custom-architecture
rose-x1
rose-medium
custom_code
Instructions to use GODELEV/Rose-Medium with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use GODELEV/Rose-Medium with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="GODELEV/Rose-Medium", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("GODELEV/Rose-Medium", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use GODELEV/Rose-Medium with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "GODELEV/Rose-Medium" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GODELEV/Rose-Medium", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/GODELEV/Rose-Medium
- SGLang
How to use GODELEV/Rose-Medium 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 "GODELEV/Rose-Medium" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GODELEV/Rose-Medium", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "GODELEV/Rose-Medium" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GODELEV/Rose-Medium", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use GODELEV/Rose-Medium with Docker Model Runner:
docker model run hf.co/GODELEV/Rose-Medium
Add hfviewer architecture card
Browse filesEmbed the hfviewer architecture visualization card in README.md.
README.md
CHANGED
|
@@ -45,6 +45,16 @@ Then five seconds later:
|
|
| 45 |
|
| 46 |
---
|
| 47 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
# What Rose-Medium Actually Is
|
| 49 |
|
| 50 |
Rose-Medium is the larger sibling of **Rose-Mini**.
|
|
@@ -311,13 +321,10 @@ Fine.
|
|
| 311 |
|
| 312 |
# Rose-Mini vs Rose-Medium
|
| 313 |
|
| 314 |
-
|
| 315 |
-
|
| 316 |
> **<img src="Comparison.png" width="1600"/>**
|
| 317 |
|
| 318 |
The important thing is that Rose-Medium does show measurable progress over Rose-Mini.
|
| 319 |
|
| 320 |
-
|
| 321 |
It is larger.
|
| 322 |
|
| 323 |
It has a longer native context.
|
|
@@ -354,7 +361,6 @@ The model has:
|
|
| 354 |
|
| 355 |
**24 layers**
|
| 356 |
|
| 357 |
-
|
| 358 |
And yet some of the reasoning scores remain fairly modest.
|
| 359 |
|
| 360 |
That makes me wonder whether the model's capacity is being allocated in the most useful way.
|
|
@@ -530,7 +536,6 @@ That's the scientific method, right?
|
|
| 530 |
|
| 531 |
# Closing Note
|
| 532 |
|
| 533 |
-
|
| 534 |
Rose-Medium is not the end of Rose X1.
|
| 535 |
|
| 536 |
If anything, it made the next step more obvious.
|
|
@@ -555,4 +560,4 @@ It taught me something.
|
|
| 555 |
|
| 556 |
And now I have to build Rose Pro.
|
| 557 |
|
| 558 |
-
**See you there.**
|
|
|
|
| 45 |
|
| 46 |
---
|
| 47 |
|
| 48 |
+
## Architecture graph
|
| 49 |
+
|
| 50 |
+
<a href="https://hfviewer.com/GODELEV/Rose-Medium?utm_source=huggingface&utm_medium=embedded_model_card&utm_campaign=GODELEV_Rose-Medium_card" target="_blank" rel="noopener">
|
| 51 |
+
<img
|
| 52 |
+
src="https://hfviewer.com/api/card.svg?source=GODELEV%2FRose-Medium&granularity=0"
|
| 53 |
+
alt="Architecture graph for GODELEV/Rose-Medium. Open in hfviewer"
|
| 54 |
+
width="100%"
|
| 55 |
+
/>
|
| 56 |
+
</a>
|
| 57 |
+
|
| 58 |
# What Rose-Medium Actually Is
|
| 59 |
|
| 60 |
Rose-Medium is the larger sibling of **Rose-Mini**.
|
|
|
|
| 321 |
|
| 322 |
# Rose-Mini vs Rose-Medium
|
| 323 |
|
|
|
|
|
|
|
| 324 |
> **<img src="Comparison.png" width="1600"/>**
|
| 325 |
|
| 326 |
The important thing is that Rose-Medium does show measurable progress over Rose-Mini.
|
| 327 |
|
|
|
|
| 328 |
It is larger.
|
| 329 |
|
| 330 |
It has a longer native context.
|
|
|
|
| 361 |
|
| 362 |
**24 layers**
|
| 363 |
|
|
|
|
| 364 |
And yet some of the reasoning scores remain fairly modest.
|
| 365 |
|
| 366 |
That makes me wonder whether the model's capacity is being allocated in the most useful way.
|
|
|
|
| 536 |
|
| 537 |
# Closing Note
|
| 538 |
|
|
|
|
| 539 |
Rose-Medium is not the end of Rose X1.
|
| 540 |
|
| 541 |
If anything, it made the next step more obvious.
|
|
|
|
| 560 |
|
| 561 |
And now I have to build Rose Pro.
|
| 562 |
|
| 563 |
+
**See you there.**
|