Instructions to use AdityaPS/SpaceLLM_v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AdityaPS/SpaceLLM_v1 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("openai/gpt-oss-20b")
model = PeftModel.from_pretrained(base_model, "AdityaPS/SpaceLLM_v1")

Transformers

How to use AdityaPS/SpaceLLM_v1 with Transformers:

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

pipe = pipeline("text-generation", model="AdityaPS/SpaceLLM_v1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("AdityaPS/SpaceLLM_v1", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AdityaPS/SpaceLLM_v1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AdityaPS/SpaceLLM_v1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AdityaPS/SpaceLLM_v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AdityaPS/SpaceLLM_v1

SGLang

How to use AdityaPS/SpaceLLM_v1 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 "AdityaPS/SpaceLLM_v1" \
    --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": "AdityaPS/SpaceLLM_v1",
		"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 "AdityaPS/SpaceLLM_v1" \
        --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": "AdityaPS/SpaceLLM_v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AdityaPS/SpaceLLM_v1 with Docker Model Runner:
```
docker model run hf.co/AdityaPS/SpaceLLM_v1
```

SpaceLLM_v1 / adapter_info.json

AdityaPS

Upload folder using huggingface_hub

e78f050 verified 23 days ago

Raw

History Blame Contribute Delete

1.97 kB

	{
	"run_id": "20260609_185647",
	"version": "v4",
	"base_model": "openai/gpt-oss-20b",
	"strategy": "LoRA on lm_head ONLY \u2014 backbone frozen \u2014 BF16",
	"key_fixes_vs_v3_fixed": [
	"lm_head weight untied (detach+clone) BEFORE get_peft_model()",
	"resize_token_embeddings tie-re-introduction guard added",
	"lora_r raised 16\u219232, lora_alpha raised 64\u2192128",
	"grad_accum raised 16\u219232, lr raised 1e-4\u21922e-4",
	"max_grad_norm tightened 0.5\u21920.3, weight_decay=0.01 added",
	"EarlyStoppingCallback added",
	"scheduler changed cosine\u2192cosine_with_restarts",
	"optimizer changed adamw_torch\u2192adamw_torch_fused"
	],
	"lora_r": 32,
	"lora_alpha": 128,
	"lora_dropout": 0.1,
	"target_modules": [
	"lm_head"
	],
	"epochs": 15,
	"learning_rate": 0.0002,
	"max_grad_norm": 0.3,
	"weight_decay": 0.01,
	"batch_size": 1,
	"gradient_accumulation": 32,
	"effective_batch_size": 32,
	"max_seq_len": 2048,
	"train_samples": 5291,
	"val_samples": 5291,
	"test_samples": 5291,
	"train_metrics": {
	"train_runtime": 24397.0057,
	"train_samples_per_second": 3.253,
	"train_steps_per_second": 0.102,
	"total_flos": 2.486936885562409e+18,
	"train_loss": 28.171284981233526,
	"epoch": 15.0
	},
	"eval_metrics": {
	"eval_loss": 1.5277799367904663,
	"eval_runtime": 412.216,
	"eval_samples_per_second": 12.836,
	"eval_steps_per_second": 12.836,
	"epoch": 15.0
	},
	"test_results_summary": {
	"mean_test_loss": 1.593661,
	"exact_match_rate": 0.0,
	"mean_token_f1": 0.270723
	},
	"final_adapter_dir": "/mnt/DATA/saurabh/aditya/SpaceLLM/fine_tuning_v2/outputs/spacellm_lora_final",
	"checkpoints_dir": "/mnt/DATA/saurabh/aditya/SpaceLLM/fine_tuning_v2/outputs/checkpoints",
	"graphs_dir": "/mnt/DATA/saurabh/aditya/SpaceLLM/fine_tuning_v2/outputs/graphs",
	"log_file": "/mnt/DATA/saurabh/aditya/SpaceLLM/fine_tuning_v2/outputs/logs/train_20260609_185647.log"
	}