Deci
/

Text Generation
Transformers
Safetensors
deci
custom_code
danaevan commited on
Commit
fd8f276
1 Parent(s): 3f282ed

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -5
README.md CHANGED
@@ -1,4 +1,3 @@
1
-
2
  # Model Card for DeciCoder-6B
3
 
4
  DeciCoder-6B is a 6 billion parameter decoder-only code completion model
@@ -19,7 +18,7 @@ Search-based technology, AutoNAC.
19
 
20
  - Google Colab [Notebook](https://colab.research.google.com/drive/1ZxG9qMlom9vn4lSGlD8PrjwHBvag94ei?usp=sharing)
21
  - Blog Post: [Introducing DeciCoder-6B: The Best Multi-Language Code Generation LLM in Its Class](https://deci.ai/blog/decicoder-6b-the-best-multi-language-code-generation-llm-in-its-class/)
22
- - Tutorial: [How to Run DeciCoder-6B on Qualcomm AI 100](https://github.com/quic/cloud-ai-sdk/tree/1.12/models/language_processing/decoder)
23
  - Questions: Feel free to contact us via our [Discord Community!](https://discord.com/invite/p9ecgRhDR8/)
24
 
25
  ## Model Architecture
@@ -61,7 +60,7 @@ refer to [https://huggingface.co/datasets/bigcode/starcoderdata](https://hugging
61
  ### Limitations
62
 
63
  The model has undergone training with source code from Python, Java,
64
- JavaScript, Ruby, RUST, C++, C, and C#. While the primary language in the source is English, it does
65
  contain other languages. Therefore, the model can produce code snippets
66
  given some context. However, there is no assurance that the resulting
67
  code will function as expected. It might be suboptimal, contain bugs, or
@@ -71,7 +70,7 @@ even exploits.
71
 
72
  Below are DeciCoder-6B's pass@1 on MultiPL HumanEval scores
73
 
74
- | Python | JavaScript | Java | C++ | C# | Rust | Go |
75
  |:----------|:----------|:----------|:----------|:----------|:----------|:----------|
76
  | 33.3% | 29.3% | 30.3% |29.93% |20.31% |20.5% |77.47% |
77
 
@@ -80,7 +79,7 @@ Below are DeciCoder-6B's pass@1 on MultiPL HumanEval scores
80
 
81
  |Inference Tool | Hardware | Prompt Length | Generation Length | Throughput (tokens/sec) |
82
  |:----------|:----------|:----------|:----------|:----------|
83
- | Qualcomm SDK | Qualcomm AI 100 | 1024 | 1024 | 531.3 |
84
 
85
  - Measured for maximal batch size on the device
86
 
 
 
1
  # Model Card for DeciCoder-6B
2
 
3
  DeciCoder-6B is a 6 billion parameter decoder-only code completion model
 
18
 
19
  - Google Colab [Notebook](https://colab.research.google.com/drive/1ZxG9qMlom9vn4lSGlD8PrjwHBvag94ei?usp=sharing)
20
  - Blog Post: [Introducing DeciCoder-6B: The Best Multi-Language Code Generation LLM in Its Class](https://deci.ai/blog/decicoder-6b-the-best-multi-language-code-generation-llm-in-its-class/)
21
+ - Tutorial: [How to Run DeciCoder-6B on Qualcomm Cloud AI 100](https://github.com/quic/cloud-ai-sdk/tree/1.12/models/language_processing/decoder)
22
  - Questions: Feel free to contact us via our [Discord Community!](https://discord.com/invite/p9ecgRhDR8/)
23
 
24
  ## Model Architecture
 
60
  ### Limitations
61
 
62
  The model has undergone training with source code from Python, Java,
63
+ JavaScript, RUST, C++, C, and C#, and Go. While the primary language in the source is English, it does
64
  contain other languages. Therefore, the model can produce code snippets
65
  given some context. However, there is no assurance that the resulting
66
  code will function as expected. It might be suboptimal, contain bugs, or
 
70
 
71
  Below are DeciCoder-6B's pass@1 on MultiPL HumanEval scores
72
 
73
+ | Python | JavaScript | Java | C++ | C# | Rust | Go |
74
  |:----------|:----------|:----------|:----------|:----------|:----------|:----------|
75
  | 33.3% | 29.3% | 30.3% |29.93% |20.31% |20.5% |77.47% |
76
 
 
79
 
80
  |Inference Tool | Hardware | Prompt Length | Generation Length | Throughput (tokens/sec) |
81
  |:----------|:----------|:----------|:----------|:----------|
82
+ | Qualcomm Cloud AI 100 SDK | Qualcomm Cloud AI 100 | 1024 | 1024 | 531.3 |
83
 
84
  - Measured for maximal batch size on the device
85