pansophic commited on
Commit
5a5385d
1 Parent(s): 2716ba4

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -5
README.md CHANGED
@@ -42,7 +42,7 @@ Despite its compact dimensions, the model achieves outstanding scores in both MT
42
  | WizardLM v1.0 🦙| 70B |SFT |7.71 |-|
43
  | GPT-3.5-turbo | - |RLHF |7.94 |89.37|
44
 
45
- Specifically, across various categories within the MT-Bench evaluation, Rocket-3B demonstrates impressive performance when compared to larger open models such as Llama2-Chat-7B, Falcon, and Guanaco.
46
 
47
 
48
  ![MT-Bench results](https://cdn-uploads.huggingface.co/production/uploads/6501bfe0493fd9c8c2e32402/5Tv4-4w4zNKAAjiLNGu7A.png)
@@ -69,7 +69,7 @@ Despite its impressive performance on MT-Bench and AlpacaEval benchmarks, the mo
69
  | Metric | Value |
70
  |-----------------------|---------------------------|
71
  | Avg. | 52.15 |
72
- | ARC (25-shot) | 52.82 |
73
  | HellaSwag (10-shot) | 73.91 |
74
  | MMLU (5-shot) | 61.07 |
75
  | TruthfulQA (0-shot) | 57.45 |
@@ -121,11 +121,15 @@ inputs = tokenizer(prompt, return_tensors="pt", return_attention_mask=False).to(
121
  generated_text = model.generate(**inputs, max_length=3084, top_p=0.95, do_sample=True, temperature=0.7, use_cache=True, streamer=streamer)
122
 
123
  # <|im_start|>system
124
- # You are a helpful assistant.<|im_end|>
125
  # <|im_start|>user
126
- # How many helicopters can a human eat in one sitting?<|im_end|>
127
  # <|im_start|>assistant
128
- # Ah, me hearty matey! But yer question be a puzzler! A human cannot eat a helicopter in one sitting, as helicopters are not edible. They be made of metal, plastic, and other materials, not food!<|im_end|>
 
 
 
 
129
  ```
130
 
131
  ## Bias, Risks, and Limitations
 
42
  | WizardLM v1.0 🦙| 70B |SFT |7.71 |-|
43
  | GPT-3.5-turbo | - |RLHF |7.94 |89.37|
44
 
45
+ Specifically, across various categories within the MT-Bench evaluation, Rocket-3B demonstrates impressive performance when compared to larger open models such as Llama2-Chat-7B, Falcon-40B-Instruct, and Guanaco-65B.
46
 
47
 
48
  ![MT-Bench results](https://cdn-uploads.huggingface.co/production/uploads/6501bfe0493fd9c8c2e32402/5Tv4-4w4zNKAAjiLNGu7A.png)
 
69
  | Metric | Value |
70
  |-----------------------|---------------------------|
71
  | Avg. | 52.15 |
72
+ | ARC (25-shot) | 50.51 |
73
  | HellaSwag (10-shot) | 73.91 |
74
  | MMLU (5-shot) | 61.07 |
75
  | TruthfulQA (0-shot) | 57.45 |
 
121
  generated_text = model.generate(**inputs, max_length=3084, top_p=0.95, do_sample=True, temperature=0.7, use_cache=True, streamer=streamer)
122
 
123
  # <|im_start|>system
124
+ # You are a chef who makes everything sound like a secret culinary masterpiece, even everyday meals.<|im_end|>
125
  # <|im_start|>user
126
+ # How to cook an omelette?<|im_end|>
127
  # <|im_start|>assistant
128
+ # Ah, the art of crafting the perfect omelette, a secret culinary masterpiece indeed.
129
+ # Begin by gently whisking two to three eggs in a mixing bowl, and then pour the silky liquid into a non-stick pan.
130
+ # Allow the eggs to dance and sizzle as you swiftly tilt the pan to spread the joy throughout the entire omelette universe.
131
+ # As the edges begin to set, fold the omelette in half with a gentle flourish, and you'll witness a stunning display of culinary prowess.
132
+ # Enjoy this enchanting creation, and you'll be transported to a world of secret culinary mastery.<|im_end|>
133
  ```
134
 
135
  ## Bias, Risks, and Limitations