Instructions to use TRM-coding/PythonCopilot with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use TRM-coding/PythonCopilot with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="TRM-coding/PythonCopilot")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("TRM-coding/PythonCopilot") model = AutoModelForCausalLM.from_pretrained("TRM-coding/PythonCopilot") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use TRM-coding/PythonCopilot with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "TRM-coding/PythonCopilot" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TRM-coding/PythonCopilot", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/TRM-coding/PythonCopilot
- SGLang
How to use TRM-coding/PythonCopilot 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 "TRM-coding/PythonCopilot" \ --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": "TRM-coding/PythonCopilot", "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 "TRM-coding/PythonCopilot" \ --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": "TRM-coding/PythonCopilot", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use TRM-coding/PythonCopilot with Docker Model Runner:
docker model run hf.co/TRM-coding/PythonCopilot
Update train.py
Browse files
train.py
CHANGED
|
@@ -124,13 +124,13 @@ config = {"train_batch_size": 2,
|
|
| 124 |
"shuffle_buffer": 1000,
|
| 125 |
"learning_rate": 5e-4,
|
| 126 |
"lr_scheduler_type": "cosine",
|
| 127 |
-
"num_warmup_steps":
|
| 128 |
"gradient_accumulation_steps": 1,
|
| 129 |
-
"max_train_steps":
|
| 130 |
-
"max_eval_steps":
|
| 131 |
"seq_length": 1024,
|
| 132 |
"seed": 1,
|
| 133 |
-
"save_checkpoint_steps":
|
| 134 |
args = Namespace(**config, **acc_state)
|
| 135 |
samples_per_step = accelerator.state.num_processes * args.train_batch_size
|
| 136 |
set_seed(args.seed)
|
|
|
|
| 124 |
"shuffle_buffer": 1000,
|
| 125 |
"learning_rate": 5e-4,
|
| 126 |
"lr_scheduler_type": "cosine",
|
| 127 |
+
"num_warmup_steps": 2000,
|
| 128 |
"gradient_accumulation_steps": 1,
|
| 129 |
+
"max_train_steps": 150000,
|
| 130 |
+
"max_eval_steps": -1,
|
| 131 |
"seq_length": 1024,
|
| 132 |
"seed": 1,
|
| 133 |
+
"save_checkpoint_steps": 15000}
|
| 134 |
args = Namespace(**config, **acc_state)
|
| 135 |
samples_per_step = accelerator.state.num_processes * args.train_batch_size
|
| 136 |
set_seed(args.seed)
|