Hugging Face's logo

bespokelabs
/
Bespoke-Stratos-32B

Text Generation
Transformers
Safetensors
English
qwen2
llama-factory
full
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
1
nlpguy commited on
Commit
268750c
·
verified ·
1 Parent(s): e457c3e

fix table

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -24,7 +24,7 @@ This model is a fine-tuned version of [Qwen/Qwen2.5-32B-Instruct](https://huggin
24
  The dataset is derived by distilling DeepSeek-R1 using the data pipeline of Berkeley NovaSky’s Sky-T1 with some modifications. More info in the dataset card at [Bespoke-Stratos-17k](https://huggingface.co/datasets/bespokelabs/Bespoke-Stratos-17k).
25
  It outperforms Qwen-2.5-32B-Instruct on reasoning benchmarks:
26
 
27
- | Metric | Bespoke-Stratos-32B | Sky-T1-32B | o1-preview | DeepSeek-R1 | DeepSeek-R1-Distill-Qwen-32B (Ours)|DeepSeek-R1-Distill-Qwen-32B (Reported)|
28
  |---|---|---|---|---|---|---|
29
  | AIME2024 | 63.3 | 43.3 | 40.0 | 79.8 | 66.7 | 72.6 |
30
  | MATH500 | 93.0 | 82.4 | 81.4 | 97.3 | 89.8 | 94.3 |
 
24
  The dataset is derived by distilling DeepSeek-R1 using the data pipeline of Berkeley NovaSky’s Sky-T1 with some modifications. More info in the dataset card at [Bespoke-Stratos-17k](https://huggingface.co/datasets/bespokelabs/Bespoke-Stratos-17k).
25
  It outperforms Qwen-2.5-32B-Instruct on reasoning benchmarks:
26
 
27
+ | Metric | Bespoke-Stratos-32B | Sky-T1-32B | o1-preview | DeepSeek-R1 | DeepSeek-R1-Distill-Qwen-32B |DeepSeek-R1-Distill-Qwen-32B (Reported)|
28
  |---|---|---|---|---|---|---|
29
  | AIME2024 | 63.3 | 43.3 | 40.0 | 79.8 | 66.7 | 72.6 |
30
  | MATH500 | 93.0 | 82.4 | 81.4 | 97.3 | 89.8 | 94.3 |