Instructions to use upiter/TinyCodeLM-150M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use upiter/TinyCodeLM-150M with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="upiter/TinyCodeLM-150M")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("upiter/TinyCodeLM-150M") model = AutoModelForCausalLM.from_pretrained("upiter/TinyCodeLM-150M") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use upiter/TinyCodeLM-150M with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "upiter/TinyCodeLM-150M" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "upiter/TinyCodeLM-150M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/upiter/TinyCodeLM-150M
- SGLang
How to use upiter/TinyCodeLM-150M 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 "upiter/TinyCodeLM-150M" \ --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": "upiter/TinyCodeLM-150M", "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 "upiter/TinyCodeLM-150M" \ --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": "upiter/TinyCodeLM-150M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use upiter/TinyCodeLM-150M with Docker Model Runner:
docker model run hf.co/upiter/TinyCodeLM-150M
Update README.md
Browse files
README.md
CHANGED
|
@@ -27,7 +27,6 @@ Despite being trained on only 72 billion tokens of text, the models outperform m
|
|
| 27 |
# Benchmarks
|
| 28 |
|
| 29 |
**Pretrained (Temperature 0)**
|
| 30 |
-
|
| 31 |
|**Benchmark**|**TinyCodeLM 150M** |**TinyCodeLM 400M** |
|
| 32 |
| :--------------------- | -----------------: | -----------------: |
|
| 33 |
| HumanEval, pass@1 | 6.1 | 6.7 |
|
|
@@ -35,7 +34,6 @@ Despite being trained on only 72 billion tokens of text, the models outperform m
|
|
| 35 |
|
| 36 |
|
| 37 |
**Edit Sequence / Instruction Tuned (Temperature-Tuned)**
|
| 38 |
-
|
| 39 |
|**Benchmark** |**TinyCodeLM 150M** |**TinyCodeLM 400M** |
|
| 40 |
| :----------- | -----------------: | -----------------: |
|
| 41 |
| HumanEval, pass@1 | 12.8 | 13.4 |
|
|
@@ -47,7 +45,7 @@ Despite being trained on only 72 billion tokens of text, the models outperform m
|
|
| 47 |
# Citation
|
| 48 |
|
| 49 |
```
|
| 50 |
-
@misc{
|
| 51 |
title={Training Language Models on Synthetic Edit Sequences Improves Code Synthesis},
|
| 52 |
author={Ulyana Piterbarg and Lerrel Pinto and Rob Fergus},
|
| 53 |
year={2024},
|
|
|
|
| 27 |
# Benchmarks
|
| 28 |
|
| 29 |
**Pretrained (Temperature 0)**
|
|
|
|
| 30 |
|**Benchmark**|**TinyCodeLM 150M** |**TinyCodeLM 400M** |
|
| 31 |
| :--------------------- | -----------------: | -----------------: |
|
| 32 |
| HumanEval, pass@1 | 6.1 | 6.7 |
|
|
|
|
| 34 |
|
| 35 |
|
| 36 |
**Edit Sequence / Instruction Tuned (Temperature-Tuned)**
|
|
|
|
| 37 |
|**Benchmark** |**TinyCodeLM 150M** |**TinyCodeLM 400M** |
|
| 38 |
| :----------- | -----------------: | -----------------: |
|
| 39 |
| HumanEval, pass@1 | 12.8 | 13.4 |
|
|
|
|
| 45 |
# Citation
|
| 46 |
|
| 47 |
```
|
| 48 |
+
@misc{piterbarg2024editseq,
|
| 49 |
title={Training Language Models on Synthetic Edit Sequences Improves Code Synthesis},
|
| 50 |
author={Ulyana Piterbarg and Lerrel Pinto and Rob Fergus},
|
| 51 |
year={2024},
|