Instructions to use jsbaicenter/JSBAI-Coder-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jsbaicenter/JSBAI-Coder-4B with Transformers:

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

pipe = pipeline("text-generation", model="jsbaicenter/JSBAI-Coder-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("jsbaicenter/JSBAI-Coder-4B")
model = AutoModelForCausalLM.from_pretrained("jsbaicenter/JSBAI-Coder-4B", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jsbaicenter/JSBAI-Coder-4B with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/jsbaicenter/JSBAI-Coder-4B

SGLang

How to use jsbaicenter/JSBAI-Coder-4B 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 "jsbaicenter/JSBAI-Coder-4B" \
    --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": "jsbaicenter/JSBAI-Coder-4B",
		"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 "jsbaicenter/JSBAI-Coder-4B" \
        --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": "jsbaicenter/JSBAI-Coder-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jsbaicenter/JSBAI-Coder-4B with Docker Model Runner:
```
docker model run hf.co/jsbaicenter/JSBAI-Coder-4B
```

jsbai-aaron commited on 24 days ago

Commit

072c219

verified ·

1 Parent(s): 31bfefa

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +10 -12

README.md CHANGED Viewed

@@ -16,9 +16,7 @@ pipeline_tag: text-generation
 # JSBAI-Coder-4B
-**An agentic coding model that runs on your laptop.** JSBAI-Coder-4B is a 4B-parameter model fine-tuned from Qwen3.5-4B for multi-step, tool-using software engineering — the capability class currently gated to 27B+ models — at a size that fits consumer GPUs.
-Built by the [James Silberrad Brown Center for AI](https://sdsu.edu) (SDSU).
 ## Results
@@ -27,16 +25,16 @@ Built by the [James Silberrad Brown Center for AI](https://sdsu.edu) (SDSU).
 | **Live-60** (real-world SWE, t32) | 15.0% | **21.7%** | **+44% relative** |
 | **IFEval** (excl-empty, 8K) | 84.66 | **87.21** | +2.55 |
 | **Held-out SWE** (unseen instances, K=8) | 10.1% | **82.9%** | **+72.8pp** |
-| MMLU-Pro (generative) | TBD | TBD | — |
-| Terminal-Bench 2.1 | TBD | TBD | — |
-*Held-out SWE: 121 never-trained coding instances solved with verified test execution — the gains are broad-spectrum generalization, not benchmark memorization. Our decontamination protocol is published with the model.*
 ## What it does
-- **Agentic software engineering**: multi-turn bug investigation, tool orchestration (bash, file editing, test running), patch generation and verification — in real repositories, inside sandboxed containers
 - **Interleaved reasoning**: thinks before each tool call, like much larger reasoning models
-- **Quality held**: instruction-following *improved* over base (87.21 vs 84.66 IFEval) — the agentic gains cost nothing on general quality
 - **Laptop-scale**: ~8GB VRAM BF16 / ~5GB NVFP4
 ## Training methodology
@@ -44,10 +42,10 @@ Built by the [James Silberrad Brown Center for AI](https://sdsu.edu) (SDSU).
 **Teacher-seeded agentic RL.** The model was produced in three stages:
 1. **Seed SFT**: a small curated set (~1,875 verified trajectories) from GLM-5.3 (a 744B-A40B open model) bootstrap the student into the agentic tool-use distribution
-2. **GRPO with verifiable rewards**: 145 batches of on-policy RL on a difficulty-banded pool of real SWE tasks (237 unique instances), where the reward is *actual test execution* — the model's patch must make the hidden tests pass
-3. **Generalization gates**: held-out never-trained instances verified the transfer at every phase boundary (10.1% → 82.9% pass rate)
-The RL phase is the capability engine: the seed SFT contributes the behavioral format, the verification-driven RL builds the problem-solving. We extensively studied on-policy distillation from the teacher (logprob-scoring) and found the signal prefers stylistic imitation over correctness at every model family and tier we tested — those negative results and the mechanism are documented in our technical report.
 ## Usage
@@ -64,7 +62,7 @@ A quantized NVFP4 variant (~5GB) and an MTP-boosted speculative decoding head (f
 - 4B-scale knowledge limits: obscure factual recall and extreme-domain reasoning remain behind larger models
 - The agent loop is tuned for sandboxed container environments; other deployment contexts are untested
-- Safety behaviors inherited from the base model; the RL phase was reward-verification only (no safety-specific training) — see the base model card
 ## Lineage & credits

 # JSBAI-Coder-4B
+**An agentic coding model that runs on your laptop.** JSBAI-Coder-4B is a 4B-parameter model fine-tuned from Qwen3.5-4B for multi-step, tool-using software engineering: the capability class currently gated to 27B+ models, at a size that fits consumer GPUs.
 ## Results
 | **Live-60** (real-world SWE, t32) | 15.0% | **21.7%** | **+44% relative** |
 | **IFEval** (excl-empty, 8K) | 84.66 | **87.21** | +2.55 |
 | **Held-out SWE** (unseen instances, K=8) | 10.1% | **82.9%** | **+72.8pp** |
+| MMLU-Pro (generative) | TBD | TBD | TBD |
+| Terminal-Bench 2.1 | TBD | TBD | TBD |
+*Held-out SWE: 121 never-trained coding instances solved with verified test execution. The gains are broad-spectrum generalization, not benchmark memorization. Our decontamination protocol is published with the model.*
 ## What it does
+- **Agentic software engineering**: multi-turn bug investigation, tool orchestration (bash, file editing, test running), patch generation and verification, in real repositories inside sandboxed containers
 - **Interleaved reasoning**: thinks before each tool call, like much larger reasoning models
+- **Quality held**: instruction-following *improved* over base (87.21 vs 84.66 IFEval). The agentic gains cost nothing on general quality.
 - **Laptop-scale**: ~8GB VRAM BF16 / ~5GB NVFP4
 ## Training methodology
 **Teacher-seeded agentic RL.** The model was produced in three stages:
 1. **Seed SFT**: a small curated set (~1,875 verified trajectories) from GLM-5.3 (a 744B-A40B open model) bootstrap the student into the agentic tool-use distribution
+2. **GRPO with verifiable rewards**: 145 batches of on-policy RL on a difficulty-banded pool of real SWE tasks (237 unique instances), where the reward is *actual test execution*. The model's patch must make the hidden tests pass.
+3. **Generalization gates**: held-out never-trained instances verified the transfer at every phase boundary (10.1% to 82.9% pass rate).
+The RL phase is the capability engine: the seed SFT contributes the behavioral format, the verification-driven RL builds the problem-solving. We extensively studied on-policy distillation from the teacher (logprob-scoring) and found the signal prefers stylistic imitation over correctness at every model family and tier we tested. Those negative results and the mechanism are documented in our technical report.
 ## Usage
 - 4B-scale knowledge limits: obscure factual recall and extreme-domain reasoning remain behind larger models
 - The agent loop is tuned for sandboxed container environments; other deployment contexts are untested
+- Safety behaviors inherited from the base model; the RL phase was reward-verification only (no safety-specific training). See the base model card
 ## Lineage & credits