Hugging Face's logo

nvidia
/
OpenCodeReasoning-Nemotron-1.1-7B

Text Generation
Transformers
Safetensors
English
qwen2
nvidia
code
conversational
text-generation-inference
Model card Files Community
wasiuddina commited on
Commit
add9b09
·
verified ·
1 Parent(s): 0a4f558

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +30 -32
README.md CHANGED
@@ -13,7 +13,7 @@ tags:
13
  pipeline_tag: text-generation
14
  ---
15
 
16
- # OpenCodeReasoning-Nemotron-7B Overview
17
 
18
  ## Description: <br>
19
  OpenCodeReasoning-Nemotron-7B-v1.1 is a large language model (LLM) which is a derivative of Qwen2.5-7B-Instruct (AKA the reference model). It is a reasoning model that is post-trained for reasoning for code generation. The model supports a context length of 64k tokens. <br>
@@ -27,37 +27,35 @@ This model is ready for commercial/non-commercial use. <br>
27
 
28
  Below results are the average of **64 evaluations** on LiveCodeBench (v5) [2408-2501].
29
 
30
- | Model | Pass@1 |
31
- |-------------------------------|--------------------|
32
- | DeepSeek-R1 | 65.6 |
33
- | QwQ-32B | 61.3 |
34
- | | |
35
- | **Distilled 7B+ Models** | |
36
- | | |
37
- | Bespoke-Stratos-7B | 14.7 |
38
- | OpenThinker-7B | 25.5 |
39
- | R1-Distill-Qwen-7B | 38.0 |
40
- | OlympicCoder-7B | 40.9 |
41
- | **OCR-Qwen-7B** | **48.5** |
42
- | **OCR-Qwen-7B-Instruct** | **51.3** |
43
- | **OCR-Qwen-7B-Instruct-v1.1** | **?** |
44
- | | |
45
- | **Distilled 14B+ Models** | |
46
- | | |
47
- | R1-Distill-Qwen-14B | 51.3 |
48
- | **OCR-Qwen-14B** | **57.7** |
49
- | **OCR-Qwen-14B-Instruct** | **59.4** |
50
- | **OCR-Qwen-14B-Instruct-v1.1**| **?** |
51
- | | |
52
- | **Distilled 32B+ Models** | |
53
- | | |
54
- | Bespoke-Stratos-32B | 30.1 |
55
- | OpenThinker-32B | 54.1 |
56
- | R1-Distill-Qwen-32B | 58.1 |
57
- | OlympicCoder-32B | 57.4 |
58
- | **OCR-Qwen-32B** | **61.8** |
59
- | **OCR-Qwen-32B-Instruct** | **61.7** |
60
- | **OCR-Qwen-32B-Instruct-v1.1**| **?** |
61
 
62
 
63
  ## Reproducing our results
 
13
  pipeline_tag: text-generation
14
  ---
15
 
16
+ # OpenCodeReasoning-Nemotron-7B-v1.1 Overview
17
 
18
  ## Description: <br>
19
  OpenCodeReasoning-Nemotron-7B-v1.1 is a large language model (LLM) which is a derivative of Qwen2.5-7B-Instruct (AKA the reference model). It is a reasoning model that is post-trained for reasoning for code generation. The model supports a context length of 64k tokens. <br>
 
27
 
28
  Below results are the average of **64 evaluations** on LiveCodeBench (v5) [2408-2501].
29
 
30
+ | Model | Pass@1 |
31
+ |----------------------------------------|--------------------|
32
+ | DeepSeek-R1-0528 | 65.6 |
33
+ | DeepSeek-R1 | 65.6 |
34
+ | QwQ-32B | 61.3 |
35
+ | | |
36
+ | **Distilled 7B+ Models** | |
37
+ | | |
38
+ | Bespoke-Stratos-7B | 14.7 |
39
+ | OpenThinker-7B | 25.5 |
40
+ | R1-Distill-Qwen-7B | 38.0 |
41
+ | OlympicCoder-7B | 40.9 |
42
+ | **OpenCodeReasoning-Nemotron-7B** | **51.3** |
43
+ | **OpenCodeReasoning-Nemotron-7B-v1.1** | **?** |
44
+ | | |
45
+ | **Distilled 14B+ Models** | |
46
+ | | |
47
+ | R1-Distill-Qwen-14B | 51.3 |
48
+ | **OpenCodeReasoning-Nemotron-14B** | **59.4** |
49
+ | **OpenCodeReasoning-Nemotron-14B-v1.1**| **?** |
50
+ | | |
51
+ | **Distilled 32B+ Models** | |
52
+ | | |
53
+ | Bespoke-Stratos-32B | 30.1 |
54
+ | OpenThinker-32B | 54.1 |
55
+ | R1-Distill-Qwen-32B | 58.1 |
56
+ | OlympicCoder-32B | 57.4 |
57
+ | **OpenCodeReasoning-Nemotron-32B** | **61.7** |
58
+ | **OpenCodeReasoning-Nemotron-32B-v1.1**| **?** |
 
 
59
 
60
 
61
  ## Reproducing our results