Hugging Face's logo

euclaise
/
Memphis-CoT-3B

Text Generation
Transformers
Safetensors
stablelm_epoch
supertrainer2000
human-data
conversational
custom_code
Model card Files Community
1
euclaise commited on
Commit
4123bc3
•
1 Parent(s): 667484a

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +4 -3
README.md CHANGED
@@ -69,11 +69,12 @@ The format for TinyCoT was:
69
 
70
  | Model | Size | Data | Method | GSM8K (5-shot) | AGIEval (English/Nous subset, acc_norm) | BIG Bench Hard (CoT, few-shot*) |
71
  |:-----------------------------------------------------------------------|--------|:--------------------|---------------|:---------------|:----------------------------------------|:------------------------------ |
72
- | [StableLM 3B Base](https://hf.co/stabilityai/stablelm-3b-4e1t) | 3B | Base | Base | 2.05% | 25.14% | 36.75% |
73
- | [StableHermes 3B](https://hf.co/cxllin/StableHermes-3b) | 3B | GPT | SFT | 3.64% | 24.31% | **37.28%** |
74
  | [MPT 7B Instruct](https://hf.co/mosaicml/mpt-7b-instruct) | **7B** | **Human**+Anthropic | SFT | 2.05% | 24.12% | 11.01% |
75
  | [OpenLLaMA 7B v2 open-instruct](http://hf.co/VMware/open-llama-7b-v2-open-instruct) | **7B** | **Human** (nearly: ecqa is an exception) | SFT | 8.64% | 23.21% | 29.84% |
76
- | [StableLM Zephyr 3B](https://hf.co/stabilityai/stablelm-zephyr-3b) | 3B | GPT | DPO | possibly contaminated (45.72%) | **33.31%** | 0.91% |
 
77
  | [**Memphis-CoT 3B**](https://hf.co/euclaise/Memphis-CoT-3B) | 3B | **Human** | Self-teaching | **18.8%** | *27.22%* | *36.92%* |
78
 
79
  *5-shot, as performed automatically by LM Evaluation Harness bbh_cot_fewshot even with num_fewshot=0
 
69
 
70
  | Model | Size | Data | Method | GSM8K (5-shot) | AGIEval (English/Nous subset, acc_norm) | BIG Bench Hard (CoT, few-shot*) |
71
  |:-----------------------------------------------------------------------|--------|:--------------------|---------------|:---------------|:----------------------------------------|:------------------------------ |
72
+ | [StableLM 3B Base](https://hf.co/stabilityai/stablelm-3b-4e1t) | 3B | Base | Base | 2.05% | 25.14% | 36.75% |
73
+ | [StableHermes 3B](https://hf.co/cxllin/StableHermes-3b) | 3B | GPT | SFT | 3.64% | 24.31% | **37.28%** |
74
  | [MPT 7B Instruct](https://hf.co/mosaicml/mpt-7b-instruct) | **7B** | **Human**+Anthropic | SFT | 2.05% | 24.12% | 11.01% |
75
  | [OpenLLaMA 7B v2 open-instruct](http://hf.co/VMware/open-llama-7b-v2-open-instruct) | **7B** | **Human** (nearly: ecqa is an exception) | SFT | 8.64% | 23.21% | 29.84% |
76
+ | [StableLM Zephyr 3B](https://hf.co/stabilityai/stablelm-zephyr-3b) | 3B | GPT | DPO | possibly contaminated (45.72%) | **33.31%** | 0.91% |
77
+ | [LIMA LLaMA 2 7B](https://huggingface.co/heegyu/LIMA2-7b-hf) | **7B** | **Human** | SFT | 4.55% | 24.55% | 36.29% |
78
  | [**Memphis-CoT 3B**](https://hf.co/euclaise/Memphis-CoT-3B) | 3B | **Human** | Self-teaching | **18.8%** | *27.22%* | *36.92%* |
79
 
80
  *5-shot, as performed automatically by LM Evaluation Harness bbh_cot_fewshot even with num_fewshot=0